Coastline Change using Time Series Satellite Images in the North Coast of Java: Study Case Indramayu Regency, West Java, Indonesia

Coastal areas have continuous change due to the interaction between sea and land, which generally causes abrasion and accretion. Indramayu Regency has a coastline of around 147 km. Indramayu Coastal has the characteristics of a sandy beach with a dominant abrasion process triggered by digging beach sand and changes in the function of mangrove land into ponds. We applied remote sensing observation to obtain coastline data on a regular basis. Currently, satellite images support the availability of Landsat data in a relatively short time and cover a wide area. Landsat imagery has been used in the last ten years to detect coastline changes in study areas based on the BILKO and Digital Shoreline Analysis System (DSAS) methods. Abrasion in 2012 – 2022 covers an area of 805.71 ha and accretion covers an area of 941.63 ha. The widest abrasion is in Pasekan with an area of 266,451 ha. The abrasion process in Pasekan is very high due to the influence of high currents and waves from the sea, and the accretion process in Cantigi is very high due to the high supply from the Cimanuk River and the many mangroves that collect sediment that enters the sea.


Introduction
The coastal area is the area that connects land with water.Coastal regions experience continuous changes due to the interaction between sea and land.The waves and wind along the coast cause constant abrasion of the rocks.This causes environmental pollution, coastal intrusion, and changes in coastline patterns.The northern coastal area of West Java has a lowland with several bays and headlands, has a curved coastline with a high accretion rate [1].The coastline is the meeting line between land and sea water which is not fixed and can change or move according to the tides and coastal erosion that occurs [2].
Indramayu Regency is located on the north coast of Java Island, consisting of 9 sub-districts that are bordered by the sea with a coastline length of 147 km [3].Indramayu Regency is also one of the districts that depend on the regional economy by utilizing the coastal area.According to data reported by Bapeda-JABAR in 2007, abrasion in the Indramayu coastal area was around 47.87% of the total length of the IOP Publishing doi:10.1088/1755-1315/1245/1/012038 2 North coast of West Java, while the stable rate was around 5.75%.This greatly affects the extent of coastal areas and communities [1].
Data is needed to see changes in the coastline and beach materials, which can be obtained periodically with the rapid development of remote sensing technology, especially satellite imagery.Satellite imagery can support the availability of data on the characteristics of an area in a relatively short time and cover a wide area so that it can provide the information needed for monitoring.The use of remote sensing data is moderately more effective than field surveys which require a long time and are difficult to monitor [4].The aim of this research to analyze changes in the coastline that occurred in Indramayu Regency from 2012 -2022.This can help determine which sub-districts experience abrasion and accretion based on the period.
Coastline changes can be described using spatial data in the form of satellite imagery.Satellite images with high resolution can illustrate coastline change in more detail and with the temporal resolution of satellite it can detect these changes easily.Usage of remote sensing data is very helpful to determine coastline changes, we can calculate the change by temporal resolution mean that we will get the information from other years before and we can study large area using this method.The use of satellite images with medium spatial resolution such as Landsat is suitable for spatial dynamics monitoring of coastline [5].
Based on previous study [6], the coastline in Indramayu Regency, between the 1990s and 2019, experienced changes in both abrasion and accretion (sedimentation), which could harm people living on the coast.Also, sedimentation causes river estuaries silting and land conflicts such as the struggle for land ownership.Furthermore, abrasion is very detrimental to the community because it damages ponds and buildings.The high level of abrasion in this Regency, especially between the 1989-2009 period, destroyed and damaged agricultural lands, pond areas, and buildings.Therefore, the government and society are making efforts to minimize abrasion.In fact, the government is building a wave barrier, both large river stones and concrete.This is considered quite effective, because between the 2009-2019 period, the level of abrasion decreased.These efforts not only reduce and minimize abrasion but also cause accretion.The accretion that appears in the coastal area is utilized by the community in various fields, including aquaculture land such as shrimp and milkfish ponds, as well as tourism.Meanwhile, accretions that appear in areas where accessibility is difficult are used for fishponds, while areas with easy access are used as tourist attractions.
The research location is in the Indramayu Regency area, it comprises 9 sub-districts stretching from East to West, including Krangkeng, Karangampel, Balongan, Indramayu, Pasekan, Cantigi, Kandanghaur, Patrol and Sukra.The aim of this research is to determine changes in the coastline in the study area in 2012 -2022 using satellite imagery data.Figure 1 showed research location in Indramayu Regency.

Figure 1.
The Indramayu coastal is in Indramayu Regency and the black line is the coastline that is the focus of this research.

Data Collection
The data used in the research process is Landsat 7 ETM+ and Landsat 8 OLI/TIRS data from 2012 to 2022, downloaded from the usgs.govwebsite.Landsat 7 has a radiometric resolution of 8 bits or 256 digital values (0-255), the spatial resolution of ETM imagery is 30 meters (channels 1-5 and 7), thermal data has a spatial resolution of 60 meters (channel 6), and panchromatic data (channel 8) with a spatial resolution of 15 meters.Image utilization is carried out for various fields, such as land and water management, pollution monitoring, detection of land surface changes, and mapping of coastline changes.Landsat 8 has an On-Board Operational Land Imager (OLI) and Thermal Infrared Sensor (TIRS) sensor with a total of 11 channels, 9 channels (bands 1-9) on OLI and 2 channels (bands 10 and 11) on TIRS.The OLI sensor has 1 near-infrared channel and 7 visible channels, which include wavelengths reflected by objects on the earth's surface, with a spatial resolution of 30 meters.The TIRS sensor has 2 thermal bands, providing a more accurate surface temperature.Image data selection is because both Landsat 7 and Landsat 8 have Near Infrared (NIR) and Shortwave Infrared (SWIR) bands which can be used to distinguish the content of groundwater and vegetation, making it easier to draw coastlines.Image data is taken from the time of tide from the study area [7].The data uses 11 satellite image data with 10 years to detect coastline changes at the study site.It is sufficiently detailed to see coastline changes and the rate of abrasion and accretion in the study area.This study uses GIS software to describe the coastline at the study area.Table 1 showed Landsat satellite sensor type and acquisition date.
Table 1.Description of Landsat Satellite image data.

Temporal Coastline Change Detection
Before analysis or identifying the coastline at the research location, the downloaded image is first prepared for optimal results.The corrections made are radiometric corrections that are processed in GIS software.Radiometric corrections function to correct pixel values so that they match what they should be by considering the factor of atmospheric disturbances as the primary source of error [8].Radiometric correction improves image quality due to reflection errors on the earth's surface and other factors, such as the direction of sunlight, weather conditions, and atmospheric conditions, so that the resulting information will be more accurate.Radiometric correction is useful for analysis multitemporal and multi-sensor data used for continuous interpretation and detection of changes.The resulting image will be sharper than the image before it was corrected.Next is an atmospheric correction which functions to correct atmospheric disturbances such as fog, dust, or smoke, which can cause bias and reflection effects on objects.Atmospheric correction was carried out using the Fast Line of Sight Atmospheric Analysis of Spectral Hypercubes (FLAASH) method.The FLAASH method is the reflectance received by the sensor and then remodelled by adding factors that affect the reflectance received by the sensor, such as the reflection of surrounding objects (adjacency effect) and the reflectance scattered by the atmosphere [6].Atmospheric correction can remove noise from the atmosphere, optimizing the image in the analysis process.
In delineating the coastline, this study uses the BILKO method to determine the boundary between land and sea by utilizing the Brightness Value (BV).The required band is the infrared band [9] Where, N is the minimum BV value for Landsat imagery, Landsat 7 (30) and Landsat 8 (7000), Ld is the NIR or SWIR band (Band 4 for Landsat 7 and Band 5 for Landsat 8) The following are the before and after results of the BILKO method in the study area to obtain the coastline every year.Visually, after BILKO method, the resulting image will be brighter, and the boundary between land and sea will look clearer (See fig 2).The following analysis is the Digital Shoreline Analysis System (DSAS) which function to calculate the change in coastline position based on time statistically and geospatially based.This method uses a point as a measurement reference, resulting from the intersection of the transects made by the researcher with the coastline based on time.The DSAS method consists of several analysis, such as [10]; (1) Shoreline Change Envelope (SCE) is a total measurement of coastline changes considering all existing coastline positions, (2) Net Shoreline Movement (NSM) is a measurement of the distance between coastline changes.Results that are positive (+) mean the coastline is advancing, and if it is negative (-), it means the coastline is retreating, (3) End Point Rate (EPR) calculates the rate of change of coastline by dividing the distance between one coastline and another.Data with positive values (+) experience accretion, and data with negative values (-) experience abrasion, and (4) Linear Regression Rate (LRR), is a statistical analysis of the rate of change using linear regression, which can be determined using the least-square regression line for all points of intersection of the coastline with the transect.

Result
The data used in the study used Landsat 7 ETM+ and Landsat 8 OLI/TIRS from 2012 to 2022.The figure below shows the BILKO method used in drawing coastlines.The results of this analysis show the difference between land and sea.Visually, the sea will appear brighter than the land.Whereas before the analysis using sea and land, BILKO looks dim.The following shows the before and after images using the BILKO method in with drawing the coastline every year.From the results of the Digital Shoreline Analysis System (DSAS), the abrasion process dominates compared to the accretion process, with an abrasion area of 805,71 ha and an accretion area of 941,63 ha (See fig 3).The DSAS method is used to analyze coastline changes to see which areas have high to low intensities of abrasion and accretion can be seen by classification [9] in Table 2.The transects used have intervals of 100 m with a transect length of 35 km along the line W -E. The highest accretion rate is in the Cantigi sub-district with a value of 29,223 m/year with a movement of 290,154 meters.the highest abrasion rate is in Karangampel sub-district with a value of -12,105 m/year with a movement of -120,192 meters (table 3).

Discussion
Coastline delineation using the BILKO method produces numbers for each coastline yearly.This can occur due to factors caused by nature, for example, tides, wave size and sedimentation the elements generated by humans, for example, land conversion.Coastline processing using the Digital Shoreline Analysis System (DSAS), we can find out that the processes that occur are divided into two, namely abrasion and accretion.We can also find out the value of the rate of change and movement of the coastline and calculate the area of abrasion and accretion in each sub-district along the Coastal Indramayu.From DSAS process resulted in 473 transects stretching along the coastline, divided into 293 abrasion transects and 180 accretion transects.Abrasion in 2012 -2022 covered an area of 805,71 ha and accretion covered an area of 941,63 ha.The widest abrasion is in Pasekan covering an area of 266,451 ha.The widest accretion is in Cantigi, with an area of 595,541 ha.The difference between abrasion and accretion in the study area is minimal, which can be caused by the range of years taken in this study.The use of Landsat imagery per year for a period of 10 years is due to see changes in the coastline in more detail.The accretion process at Cantigi is very high because the high sediment supply from the Cimanuk River and the many mangroves that hold the sediment that enters the sea.The abrasion process in Pasekan is very high because the high influence of currents and waves from the sea.
Based on previous study [5], coastline changes were analysed from 1989-2019, using Landsat 5 and Landsat 8 satellite imagery data, using the BILKO method and the overlay method.The results were for the value of the abrasion rate in 1989-2019 covering 2,127 ha and accretion covering an area of 746 ha.Whereas in the current study with 1 year, from 2012-2022, using the BILKO and DSAS methods, the value of the abrasion area is 805,71 ha and the accretion area 941,63 ha.There are differences in values, and this is due to the time distance used to analyse, the method used, as well as the level of accuracy of researchers in analysis.

Conclusion
The research conducted is research that focuses on changes in the coastline in Indramayu Regency in 2012-2020 using satellite imagery data.The methods used include radiometric correction, which functions to sharpen the image before analysis, atmospheric correction, which functions to remove noise from the atmosphere, which can optimize the image in the analysis process.The BILKO method is used to delineate coastlines that utilize the NIR and SWIR bands.The DSAS method is used to analyse the coastline and the rate of abrasion and accretion in the study area.From the results of the Digital Shoreline Analysis System (DSAS), the abrasion process dominates compared to the accretion process, with an abrasion area of 805,71 ha and an accretion area of 941,63 ha.The widest abrasion is in Pasekan covering an area of 266,451 ha.The widest accretion is in Cantigi, with an area of 595,541 ha.The difference between abrasion and accretion in the study area is minimal, which can be caused by the range of years taken in this study.The use of Landsat imagery per year for a period of 10 years is due to see changes in the coastline in more detail.The accretion process at Cantigi is very high because the high sediment supply from the Cimanuk River and the many mangroves that hold the sediment that enters the sea.The abrasion process in Pasekan is very high because the high influence of currents and waves from the sea.

Figure 4 .
Figure 4. Map of abrasion and accretion of Indramayu Regency.The red line indicates the abrasion area, and the blue line indicates the accretion area.

Figure 5 .
Figure 5. (A) The abrasion shift in Indramayu district is 306,91 meters and the accretion shift is 1588,25 meters.(B) The abrasion rate in Indramayu district is 30,91 m/year and the accretion rate is 159,96 m/year.
; on Landsat 7 required band 4, and on Landsat 8 required band 5. Infrared waves have a high reflectance to reflectance on water.The BV value is required in the separation of land and sea.Meanwhile, the BILKO formula used is as follows:

Table 2 .
Classification for coastline change (EPR and LRR values).

Table 3 .
Rate and change of coastline for each sub-district along the Indramayu coastal.