Mangrove Density Mapping from Landsat 8/9 OLI Imagery in Dompak Island, Indonesia: A Study from 2017 to 2022

. Dompak Island, a small island in the Riau Islands Province, has experienced massive urban development by local governments over the last decade. This activity has impacted the sustainability of vegetation and mangrove forests on Dompak Island. A decrease in the mangrove area can affect the condition and sustainability of the species. This study aimed to analyze the changes in mangrove density using Landsat 8/9 OLI satellite imagery of Dompak Island, Tanjungpinang City, from 2017 to 2022. This study used an unsupervised classification method, which used the Normalized Difference Vegetation Index (NDVI) to identify the changes and density of mangroves, and an algorithm to obtain the greenness value of vegetation objects. This study found changes in the area and growth of mangroves. There is a decrease in mangrove density globally. The level of rare density in 2017 and 2022 was still at 23%, the medium density increased by 8%, and the high density decreased by 9%. Based on mapping results, the area of mangrove forests has decreased by 4 ha from 2017 to 2022, namely from 93 ha to 89 ha..


Introduction
The concept of "blue carbon" has gained prominence in the global discourse on climate change because it pertains to the carbon sequestration potential of coastal and marine environments, including mangrove forests [1].These unique ecosystems can sequester carbon at rates significantly higher than those of the terrestrial forests.Mangrove forests act as a vital carbon sink by trapping CO2 in both their biomass and the sediment below, making them an integral part of strategies to combat rising CO2 levels.
In the mangrove ecosystem, some organisms interact with environmental and other factors [1,2].Mangrove forests have several functions, one of which is ecological, such as retaining mud, reducing waves, and protecting beaches from abrasion.The higher the percentage of mangrove forest density, the higher the ability to reduce waves [3].The density index can be assessed using the Normalized Difference Vegetation Index (NDVI) with remote sensing method [4], which characterises the amount of greenness of vegetation and then calculates the near-infrared and red reflected by plants to identify the density of vegetation as land with no vegetation or that is not vegetated [5][6][7].
Monitoring changes in mangrove forest density is essential to understanding the impacts of climate change and urban development on these vital ecosystems.Remote sensing, using advanced technology like Landsat 8 and 9 imageries, helps keep an eye on these places [8].By looking at plant data from far away, we can learn about how healthy and big these areas are.This helps us make smart choices for conservation and restoration [9].Dompak Island in Riau Island Province is rich in mangrove plants [10].As much as 27.6% of all mangrove vegetation in Tanjungpinang City is on Dompak Island [11].This area experiencing massive urban development by the local government, which has resulted in large-scale felling of forests, which are then converted into office areas and housing, as well as the construction of infrastructure and roads, as well as other infrastructure [12,13].A decline in its forest area can affect the species and its ecosystem condition.According to statistics [14,15], there have been a lot of land use changes on that island, so analyzing the impact on mangroves has become important.
Many studies on mangroves on Dompak Island have been carried out previously, such as the composition of mangrove species and density [10,16], mangrove carbon stock in sediment [12], object-based mangrove land cover change (OBIA) studies [13], and the level of mangrove damage [17].However, there is no data on the latest spatial analysis of mangrove density using open-source satellite imagery such as Landsat 8 imagery.Information about vegetation obtained from remote sensing imagery primarily focuses on variations in the greenness of leaves and the spectral properties of the canopy.The most common way to validate this information is by establishing a direct or indirect connection between the acquired Vegetation Index (VI) and on-site measurements of vegetation attributes like coverage, Leaf Area Index (LAI), biomass, growth, and vitality.Another approach involves comparing VI from sentinel plants with VI from the same plant for calibration purposes using direct and geographical reference methods [18].Remote sensing analysis of vegetation often employs the vegetation index method, with one well-known index being the Normalized Difference Vegetation Index (NDVI).NDVI is highly responsive to changes in green vegetation and is linked to the physical and physiological alterations in plants [5,19].NDVI has gained popularity for its versatility, ease of acquisition [20], and strong correlation with parameters such as density, productivity, photosynthetic capacity, and leaf area index [21].Previous studies have effectively employed NDVI for land cover change detection [22][23][24].Considered to be Landsat satellite imagery, it is suitable for this research because it is in the medium resolution category and requires accuracy to determine the level of vegetation density and the area distribution of mangrove forests.This study aims to conduct a comprehensive multi-temporal assessment of changes in mangrove forest areas within Dompak Island, Tanjungpinang City -Indonesia using Landsat 8 and 9 derived vegetation indices.In this context, multi-temporal monitoring using remotely sensed data, including vegetation indices, offers in-depth insights into these changes.By examining the alterations in these ecosystems over time, this study seeks to highlight their vulnerability to climate change and human activities, emphasizing the importance of their preservation in the context of small island conditions.

Study Area
The location of this research is located on Dompak Island, Tanjungpinang City, Riau Islands Province.Astronomically, it is located at 0°51' -0°59' North Latitude and 104°23' -104°34' East Longitude (Figure 1).Additionally, in this area, there is a diverse range of mangrove species that need to be protected [10], especially with the increasing urbanization.

Data Source and Tools
The tools and data used in this study consist of GIS software, some hardware to collect field data from field surveys and Tanjung Pinang City Environmental Service Department, and secondary data from satellite Landsat Imagery in the year 2017 and 2022.It can be seen in Table 1 and Table 2.

Method
The method used in this research is a quantitative descriptive approach, which aims to explain a scenario with the help of literature studies.Figure 2 depicts the flow diagram for the operation and analysis.Field data collection (ground check) is a key activity in satellite image interpretation.This activity shows the actual ecosystem conditions in the field.The ground check activity entails correcting the distribution and density results of mangroves from satellite imagery with density data and mangrove canopy closure in the field using the Global Positioning System (GPS).

Sampling location determination
The sampling location determined by unsupervised classification method.The sampling was done at randomly and on purposively (taking into consideration how it represents each class).As documentation, the data collected includes photographs of mangroves.

Data analysis
At the shape of pra-processing analysis, the image should be entered into the QGIS application, followed by geometric corrections and map registration, which aim to update the coordinates according to the standard OSM in QGIS utilizing the RBI map.The following step is for a radiometric correction test, which attempts to convert the data in the image from a digital number into reflectance, which will be processed using the NDVI method.The NDVI approach was utilized at the data analysis step to determine the level of mangrove density.Analysis data processing carried out by transform corrected data to vegetation index value using equation 1 [20].In this case transform to NDVIs on the year 2017 and 2022.
The formula to transformation NDVI where NIR is the Near Infrared band that is band 5 with wavelength 0.85 -0.88 micrometres, and RED is the Red band that is band 4 with wavelength 0.64 -0.67 micrometres.
The calculation results using the NDVI method will be reclassified into 3 categories according to Table 3., namely seldom, medium, and high density.The following is a table assessing the level of mangrove canopy density based on NDVI values.Vegetation index calculation informed the area and distribution of mangroves each year of data on Dompak Island, Tanjungpinang City.The result of data analysis is a distribution map of mangrove changes between 2017 and 2022 to be used as management direction.These management directions are usually adapted to coastal communities, such as community socio-economics, community knowledge about mangroves, and community participation in mangrove management.Radiometric correction to determine the area and density of mangroves uses bands 4 and 5 as the main ingredients.Band 4 or red channel and band 5 or near-infrared channel are used because they follow the NDVI formula; apart from that, these two bands can identify vegetation objects with high spectral responses from the greenness value [20].Radiometric correction is used to correct images from errors such as surface reflections or the curvature of the earth, and so on, so that the information obtained is more accurate.The following compares the uncorrected image and the radiometrically corrected image in Figure 3.

The mangrove density in 2017.
After correction, both bands on Landsat 8 Imagery were analyzed using the NDVI method.
The NDVI results on the 2017 Landsat 8 OLI image are as follows Figure 4.In this research, there were 2 image cuts, namely the shp file of Dompak Island using the administrative boundaries of Tanjungpinang City and the shp file of mangrove using data from Tanjung Pinang City Environmental Service Department collected in 2022.There are 3 determinations of density classes based on the area in hectares, namely rare, medium, and high-density classes.This aims to facilitate analysis in the form of map visualization and is also based on the Department of Forestry's (2005) assessment of mangrove canopy density levels.The results of this analysis are presented in Table 4.The research results show that the total area of mangrove vegetation on Dompak Island in 2017 was 93 ha.In 2017, the mangrove area was 11 ha (12%) in rare density, 22 ha (24%) in medium density, and 60 ha (65%) in high density.The map classification of mangrove density in 2017 can be seen in Figure 5.

The Mangrove Density in 2022.
The results of NDVI analysis from Landsat 9 0LI imagery in 2022 in the Dompak Island area are shown in Figure 6.The analysis method used is the same as that used in Landsat 8 for 2017 imagery because Landsat 8 OLI and Landsat 9 OLI are the same, only with increased radiometric precision.
The research results show that the total area of mangrove vegetation on Dompak Island in 2022 was 93 ha.The result is shown in Table 5 where in 2017, the mangrove area was 11 ha (12%) in rare density, 22 ha (24%) in medium density, and 60 ha (65%) in high density.
The map classification of mangrove density in 2017 can be seen in Figure 7.

Analysis of Changes in Mangrove Area and Density in 2017 and 2022.
The mangrove ecosystem has been identified as a transition ecosystem between land and marine ecosystems, and it offers a vital function as well [25].Although it can persist in spots where the highest tides reach, so that referred to as coastal forest [26].Dompak Island, Tanjungpinang City, Riau Islands Province, is home to one of Indonesia's mangrove vegetation zones.To assess vegetation density, researchers often perform various comparative analyses.One common method involves using Landsat-8 imagery, which has a 30-meter resolution and includes red and near-infrared wavelengths that are highly sensitive to vegetation.In this study, data on mangrove density and area were obtained from Landsat 8/9 OLI images.
These images were processed using the NDVI (Normalized Difference Vegetation Index) formula, which calculates a vegetation density index based on the near-infrared and red bands.NDVI is a widely used index for monitoring plant growth and has been adapted for various applications in vegetation analysis [17].The study's findings, comparing NDVI values for the years 2017 and 2022, are presented in Table 6.
Based on the Table 6 mangrove forest density on Dompak Island in 2017, the rare density were identified as red with an NDVI value range of 0.01 to 0.42.The medium density identified in yellow in the NDVI value range of 0.42 to 0.69.Mangrove areas in high density identified in green have an NDVI value range of 0.69 to 0.87.The green color that still dominates the mangrove vegetation proves that the mangroves on Dompak Island in 2017 were considered good.The mangrove locations with the highest density are in the northwest direction of Dompak Island, which is very close to the coast and several points in the northwest area.The yellow color or medium density level tends to be in mangroves with dense density.Meanwhile, dense mangroves are rarely far from the coast, and the possibility of them not being exposed to water at low tide is very high.According to Rahmad et al., (2020), the influence of sea tides has a good impact on the adaptability of several types of mangroves.
The image used to determine the area and density of mangroves on Dompak Island in 2022 is Landsat 9 OLI 2 Image.This image was launched in 2021 and orbits at an altitude of 705km above the earth.Landsat 9 OLI 2 imagery is an update of Landsat 8 OLI imagery.
There has not been much use of Landsat 9 OLI 2 imagery, so this is an opportunity to find out its capabilities in mapping mangroves.density that are identified as green have an NDVI value range of 0.57 to 0.73.In 2022, the mangrove area located in the Southeast direction directly facing Roro Dompak Harbor will experience additional mangrove areas in several categories, namely rare, medium and high.However, the medium category dominates the area around the bridge.This is thought to be caused by the conversion of land into a restaurant and resulting in the destruction of mangrove vegetation in the area.
In Mangroves with changes from high to medium identified as turquoise blue are 18.4 ha.
Over the five-year period, it is evident that high-density mangroves saw a significant decrease, while low-density mangroves increased in terms of the percentage of total area.This may indicate changes in the mangrove ecosystem's condition on Dompak Island during this period, potentially influenced by various factors such as human activities, climate change, and others.

Management directions.
In order to manage the mangrove ecosystem on Dompak Island, several crucial steps need to be taken.Firstly, regular monitoring and research should be conducted to understand changes in mangrove density and the influencing factors.Next, habitat preservation is key, including protecting existing mangrove areas and preventing illegal activities that harm the ecosystem.Mangrove rehabilitation programs should also be implemented to restore degraded areas by planting suitable mangrove seedlings.Moreover, educating and raising awareness among the local community about the importance of mangroves and involving them in conservation efforts are essential.Sustainable management, proper waste management, and adaptation to climate change should also be part of the mangrove management strategy.Collaboration with various stakeholders such as local government, NGOs, and the local community should be enhanced to achieve sustainable mangrove conservation goals.

Conclutions
Dompak Island is one of the areas being developed to become the center of government in Tanjungpinang City, Riau Island Province.Nevertheless, the local government still

Fig. 8 .
Fig. 8.The Changes of Mangrove density Map in Dompak Island both 2017 and 2022.The changes of mangrove area and density both 2017 and 2022 shown in figure 8. Mangrove areas with rare density levels in 2022, identified in red, have NDVI values ranging from 0.06 to 0.42.Mangrove areas with medium density levels identified as yellow have NDVI values ranging from 0.42 to 0.57.Meanwhile, mangroves with a high level of

Table 1 .
The research data and usability.

Table 2 .
The research tools

Table 3 .
The canopy density class The image used to determine the area and density of mangroves on Dompak Island in 2017 was Landsat 8 OLI imagery and in 2022 was Landsat 9 OLI Imagery.It was acquired on September 26, 2017 in the name LC08_L1TP_125059_20170926_20200903_02_T1, and the another in the name LC09_L2SP_125059_20220425_20230418_02_T1 was acquired on April 25, 2022.The imageries processed to obtain radiometric correction results.

Table 4 .
Results of NDVI Analysis of Mangrove Area and Density in 2017.

Table 5 .
Results of NDVI Analysis of Mangrove Area and Density in 2022.

Table 6 .
Comparison results of NDVI values for 2017 and 2022.
2022, mangrove vegetation has been spreading across the North, West, East, Southeast, and South areas of Dompak Island.Detections of mangrove changes in 2017 and 2022 are spread across the North to the North West, North to North East, Southeast, and South of Dompak Island.The area with cream color in 2017 was declared non-existent, and then in 2022 it was declared rare with an area of 5.98 ha.The area with a dark brown color is a transition area from none to medium density, covering an area of 11.4 ha.The area colored pink in 2017 was categorized as non-existent but will change to dense in 2022, covering an area of 5.96 hectares.The area colored blue in 2017 and 2022 is a permanent area, namely Rarely, with an area of 0.57 ha.Mangroves with changes from rare to moderate are identified with a blue-gray color of 4 ha.The detection of changes from the rare to high density category marked in maroon is 0.56 ha.The area colored light purple, which was categorized as medium in 2017 but changed to missing in 2022, experienced an area change of 6.7 ha.The mangrove category remains medium-medium, with a light brown color covering an area of 9.2 ha.Mangroves that were categorized as medium and changed to a rarely colored dark purple experienced a change of 4.48 ha.Detection of changes with a moderate to dense transition with a dark green color covering an area of 1.8 ha, The area in the pink color in 2017 was categorized as high but lost in 2022, covering an area of 11.1 ha.Mangroves with a change from high to rare identified by the color orange are 26.3 ha.