Estimation of vegetation density using NDVI and carbon stock in the Utilization Zone of Gunung Halimun Salak National Park

Forestry and Other Land Uses (FOLU) is one of the critical sectors in low carbon and climate resilience development. Gunung Halimun Salak National Park (GHSNP) can potentially mitigate climate change by increasing carbon stocks. This study aimed to estimate vegetation density using the Normalized Difference Vegetation Index (NDVI), estimate biomass and aboveground carbon stocks, and analyze the relationships of NDVI with vegetation diversity and carbon stocks in the utilization zone of GHSNP. The method used was vegetation analysis and carbon estimation using allometric and destructive testing for seedling and understory levels. The total plots were 15 (a plot size of 50 × 50 m). NDVI was categorized into three classes, specifically class 1 (0,147 - 0,276), class 2 (0,276 - 0,321), and class 3 (0,321 - 0,493). The utilization zone of GHSNP consists of 56 plant species from 40 families. The research location was dominated by Pinus merkusii. The average biomass and carbon concentrations were 131,14 Mg/ha and 61,64 Mg C/ha, respectively. Carbon concentration strongly correlated with the basal area (r = 95,2%). The NDVI value strongly correlated with the number of individual trees per hectare. NDVI can be used as an approach to estimating forest cover.


Introduction
Climate change and global warming (global warming) are current global forestry issues.The increase in the concentration of greenhouse gases and carbon emissions are the main factors causing climate change.Indonesia ratified the Paris Agreement to prevent a global temperature rise of 2 ˚C and limit global temperature rise to 1,5 ˚C by reducing greenhouse gas emissions.Forestry and Other Land Uses (FOLU) is one of the important sectors in the long-term strategy of low carbon development and climate resilience.FOLU Net Sink 2030 sets a target for Indonesia to achieve a level of carbon sequestration equal to or higher than the level of emissions.This commitment is contained in the Nationally Determined Contribution (NDC) document submitted to the United Nations Framework Convention and Climate Change (UNFCCC).The NDC document sets Indonesia's emission reduction target of 29% on its own and 41% with international support in 2030 [1].
Gunung Halimun Salak National Park (GHSNP) can potentially mitigate climate change by increasing carbon stocks.GHSNP is Java's largest nature conservation area, with high species diversity.Some have experienced a decline in population numbers due to forest degradation in utilization zones 1315 (2024) 012041 IOP Publishing doi:10.1088/1755-1315/1315/1/012041 2 close to human activities.This phenomenon disrupts ecosystem services, as well as the opening of the forest canopy.
Monitoring forest canopy and vegetation density can be done through remote sensing using the Normalized Difference Vegetation Index (NDVI) approach.NDVI describes the level of greenness of a plant through the spectral reflection of green leaves as an indicator of the presence of vegetation and forest quality [2].Therefore, further research is needed to determine the contribution of GHSNP to climate change mitigation actions using a spatial approach.This study aimed to estimate the vegetation density using NDVI, estimate biomass and aboveground carbon stocks, and analyze the relationships of NDVI with vegetation diversity and carbon stocks in the utilization zone of GHSNP.This research is expected to provide information to GHSNP regarding vegetation density, species diversity, and carbon stocks to be used as a basis for decision-making for GHSNP officers in managing the GHSNP area.The research results are expected to provide information regarding the contribution of the GHSNP in mitigation actions toward Indonesia's FOLU Net Sink 2030.

Study Site and Period
This study was conducted in the utilization zone of the Gunung Salak I Resort, Gunung Halimun Salak National Park (6°40'16.8"S106°44'48.5"E).The utilization zone at Gunung Salak I Resort has an area of 247,45 hectares.However, this research was conducted specifically in Tamansari Village and Sukamantri Village, Bogor Regency, West Java Province, Indonesia.The vegetation density, biodiversity, and carbon stock were analyzed from October to December 2022.The map of the study site is presented in Figure 1.The utilization zone is one of the zones in the Gunung Halimun Salak National Park area.The utilization zone is designated to fulfill the functions of utilization, environmental services, and ecotourism in the GHSNP area.For example, hiking trails for visitors are included in the utilization zone of GHSNP.

Tools and Materials
The tools used in this study were: GPS, Avenza Maps, and compass for land navigation; measuring tape, diameter tape, haga hypsometer, and tally sheet for vegetation analysis; camera for documentation; trash bag, zip-lock, digital scales, oven, and A3 envelopes for destructive sampling purpose; Microsoft Office Word, Microsoft Office Excel, ArcGis, and SPSS for data analysis.Spatial analysis in this study used ArcGIS 10.8.2, which is the current release of ArcGIS Desktop.The materials in this study were vegetation in the utilization zone of Tamansari Village and Sukamantri Village, Gunung Salak 1 Resort, GHSNP.Vegetation density classification is used the Natural Breaks (Jenks) method in ArcGIS 10.8 [3].

Plot establishment.
Single plots were constructed using stratified random sampling.The number of sample plots used is five plots for each density class.Plots measuring 50 m × 50 m were used for tree vegetation analysis and sub-plots were used for poles (10 m × 10 m), saplings (5 m × 5 m), and seedlings and understorey (2 m × 2 m).The plot is made with a size of 50 m × 50 m so that the data obtained is more representative of the area of the plot, which is smaller than the pixel size of Landsat 8 imagery (30 m × 30 m in multispectral mode) [4].

Species Composition and Structure of Tree Stands
Measurement.Understorey and seedlings (saplings starting to germinate to a height of <1.5 m) were recorded with species names and the number of individuals of each.Vegetation at the level of saplings (≥1.5 m in height and <10 cm in DBH), poles (10 -20 cm in DBH), and trees (≥20 cm in DBH) was measured for total height, branch-free height, and DBH (diameter at breast height) on each individual.

Aboveground Biomass and Carbon Stock Estimation.
Biomass and carbon stock estimation was focused on the aboveground.The destructive method was used to estimate the understorey and seedling biomass.All understorey plants and seedlings in the 2 m × 2 m sub-plot were harvested to measure total wet weight without the root.Sub-samples (200 -300 grams from total wet weight) were heated in the oven at 80 o C for 48 hours [5].The weight of the sample that has been heated is the dry weight.The total dry biomass of understorey and seedlings can be calculated using the Equation 1.
Total dry weight (kg/m 2 ) = Total fresh weight (kg) × Subsample dry weight (g) Subsample fresh weight (g) × Sample area (m 2 ) (1) Stand biomass was measured using a non-destructive method.The non-destructive method was used because the stands measured were located in the National Park area, making it impossible to use destructive methods.Total height and DBH were measured at the growth rate of saplings, poles, and trees to predict biomass through an allometric equation.Pinus merkusii stands dominate the utilization zone of Gunung Halimun Salak National Park, so the biomass of P. merkusii and other tree species (mixed species) is calculated using the equation in Table 1.
Carbon stock (Mg/ha) = Biomass (Mg/ha) × 0,47 provide an overview of the influence or role of a type of vegetation in a community.Density, frequency, and dominance can be calculated using the following Equation 3 to Equation 8 [9].

Density (D) =
Total number of individuals of a species found Total area examined Understorey, seedlings, and sapling IVI (%) = RD + RF (9) Pole and tree IVI (%) = RD + RF + RDo (10) The level of species diversity can be seen from the value of the species biodiversity index.The Shannon-Wiener species diversity index (H') describes an ecosystem's species diversity, stability, and maturity level, with the index value is around 1,5 to 3,5 and rarely exceeds 4 [10].The species dominance index (C) shows the pattern of concentration and dominance of a species in a plant community with the values 0 -1 [11].The Margalef species richness index (R) shows a community's species richness.An ecosystem's species richness can be considered high if it has a value of more than 2,05 [12].The species evenness index (E) shows a community's relationship between abundance and species diversity.The relatively even species abundance is indicated by an E value close to 1 [13].

The Relationships Between NDVI and the Number of Species, Tree Density, Species Diversity,
Basal Area, and Carbon Stock.The classical assumption test was carried out before testing the hypothesis to ensure that the regression model equation was said to be good [14].The classic assumption test was carried out using the normality and heteroscedasticity tests.After that, a correlation and regression test was carried out to measure the relationship between all the variables.Then, the regression model was tested for validity test to show the accuracy of the selected prediction model.

Vegetation Density Using Normalized Vegetation Index (NDVI) in the GHSNP Utilization Zone
Normalized Difference Vegetation Index (NDVI) analysis result was categorized into three classes.The results of the NDVI analysis are shown in Figure 2. One class of vegetation density is described by one color.Areas with a high level of vegetation density are shown in a darker color (green) due to the high reflection of light produced by the vegetation [15].NDVI values are divided based on the Natural Breaks (Jenks) classification method in ArcGIS 10.8.This classification method can group data based on data distribution and the largest range value.NDVI values and colors in each class are presented in Table 2. Based on Table 2, the value range is above 0, so the three classes are considered vegetated areas.Each vegetation density class has different vegetation cover conditions.The general condition of the vegetation cover for each class in the GHSNP utilization zone is presented in Figure 3.

Vegetation Composition
Vegetation analysis conducted on 15 plots with a total area of 3,75 ha showed 56 plant species from 40 families.The most common families found were Arecaceae and Urticaceae, with 4 species each, and Euphorbiaceae, with 3 species.The number of species found is presented in Table 3.
Table 3.The number of species found in the utilization zone of GHSNP.Based on Table 4, the understorey with the highest IVI is Oplismenus hirtellus and Piper aduncum.
Calliandra calothyrsus dominates all vegetation density classes at the seedling and sapling levels because it is a fast-growing species.GHSNP utilization zone, where most of the area is in the lowland rainforest zone, has undergone many changes and degradations.The high IVI value of Homalanthus populneus and Macaranga sp. in each class indicates forest damage [16].
The utilization zone of GHSNP is dominated by Pinus merkusii, which were planted in monoculture by Perum Perhutani before the area changed into a national park.That is shown by the high IVI values of pine in all density classes.P. merkusii natural regeneration is not found in all plots.This phenomenon can occur because P. merkusii needle litter is difficult to decompose due to the high lignin content, so fallen pine seeds cannot reach the ground [17].It also makes it difficult for other species to grow in pine-dominated areas.

Vegetation Diversity Index
The level of species diversity can be seen from the value of the species diversity index (H'), species dominance index (C), species richness index (R), and species evenness index (E).The values of H', R, C, and E in the utilization zone of Gunung Salak I Resort are presented in Table 5.

Table 5. Result of vegetation index.
Growth level Based on Table 5, the highest H' value is shown by understorey in class 1 (2,35), while the lowest H' value is shown by the growth rate of trees in class 1 (0,05).The results indicate that the research location cannot be stable against disturbances [18].Plant communities with high species diversity will have a low dominance value (close to 0).The highest C value is owned by the growth rate of trees in class 1.It is related to the low value of the index H' because one species dominated the observation site.The Margalef species richness index (R) shows a community's species richness.Table 7 shows the growth rate of trees in all density classes included in the low category.The research location is monoculture planted with P. merkusii.Species evenness index (E) shows a community's relationship between abundance and species diversity.Evenness values will be low (close to 0) if there are species that dominate.The lowest E value is found at the tree level in class 1, comparable to the H' and R values.A low E value indicates that the condition of the research plot only supports a few species to grow and dominate the area [19].

Stand Structure
The horizontal vegetation structure describes vegetation's availability level at each growth level.An overview of the horizontal structure of vegetation that relates the density of individual trees per hectare based on their diameter class can be seen in Figure 4. Figure 4 shows that the resulting horizontal vegetation structure curve forms an inverted "J".That indicates that the distribution of regeneration shown can guarantee the sustainability of the forest.Tree diameter classes that vary are influenced by differences in tree age, genetic factors for each species, quality of growing sites, and competition for nutrients and sunlight.Tree regeneration in the early growth phase requires much energy, giving rise to natural selection.This phenomenon reduces the number of individuals that survive at each increase in diameter class [20].

Estimation of Biomass and Carbon Stock
Biomass is the total mass of material derived from living things, including organic and dead organic matter.The results of estimating aboveground biomass and carbon stocks in the GHSNP utilization zone are presented in Table 6.Table 6.Aboveground biomass and carbon stocks.Table 6 shows that the estimated value of biomass and carbon stock increases with increasing vegetation density.The highest biomass and carbon stocks were shown by class 3, namely 141,64 Mg/ha and 66,56 Mg C/ha, respectively.Aboveground biomass greatly contributes to total biomass, equal to 80,1%.Aboveground biomass was distributed into stems, branches, leaves, and stalks of 55,5%, 17,4%, 4,9%, and 2,4 %, respectively.It shows that the stem will always have the largest proportion of biomass affected by its diameter.Trees have a percentage of 85,16%, while the contribution percentage of understorey and seedlings is below 1% [21].
Forest carbon stored in biomass can be used to describe the condition of forest ecosystems.The potential of biomass and carbon stocks in the GHSNP utilization zone has an average value of 131,14 Mg/ha and 61.64 Mg C/ha.The results are similar to research in the GHSNP traditional zone, which is directly adjacent to the TNGHS utilization zone, namely a biomass of 108.55 Mg/ha and a carbon concentration of 51,02 Mg C/ha [2].

The Relationships Between NDVI, Stand Density, Tree Diversity, and Carbon Stock
The results of the classical assumption test that was carried out before testing the hypothesis showed that all data were normally distributed and did not show symptoms of heteroscedasticity.That indicates that correlation and regression analysis can be performed.Correlation test results Pearson presented in Table 7.The correlation test Pearson carried out aims to measure the closeness of the relationship between the independent variable and the dependent variable.Based on Table 7, the strongest variable relationship is shown by the correlation of carbon concentration with LBDS, which equals 0,952.As much as 95,2% of the increase in carbon concentration is affected by LBDS.It follows the [21] statement that as much as 55,5% of the standing biomass is stored in the stem, which is influenced by the diameter size.
Based on the decision-making assumptions in the correlation analysis, the correlation test results between NDVI and all dependent variables are uncorrelated.The NDVI value has the highest correlation with the number of individuals (0,390), while the density strongly correlates with carbon concentration (0,772).That indicates NDVI correlates with carbon concentration indirectly through the number of species.These results are influenced by the time the satellite image was taken, carried out in March, and field data collection in October.These different periods allow for changes in vegetation conditions in the field.In addition, old trees that dominate the field are no longer active in physiological processes, so the leaves reflect little red light.Pinus merkusii, which dominates the land, also has a small leaf area, so the canopy density that covers the land tends to be low.The regression test was carried out to ascertain and know how far the NDVI value (independent variable) can predict the dependent variable (Figure 5).The regression analysis results show that the coefficient of determination obtained between NDVI and all dependent variables is classified as very weak, namely 0,04%, 17,06%, 0,29%, 4,01%, and 2,52%, respectively.The coefficient of determination in the 0% -19,9% value range is classified as very weak [22].
The estimation model is validated to see and know the accuracy of the resulting model.The results of testing the validity of the five-parameter estimation model using NDVI are presented in Table 8.

Conclusion 4.1. Conclusion
NDVI analysis in the utilization zone of GHSNP shows three classes of vegetation density.There are 56 species of plants from 40 families in the GHSNP utilization zone.The horizontal structure of the vegetation forms an inverted "J" curve, which means that the distribution of individuals can guarantee the sustainability of the forest due to the availability of regeneration.The average biomass and carbon concentrations obtained were 131,14 Mg/ha and 61,64 Mg C/ha, respectively.Carbon concentration correlates highest with LBDS (r = 95.2%).The NDVI value correlates with the density.The study area was dominated by old trees that no longer active in physiological processes.Therefore, the leaves reflect little red light which made the NDVI values show the land has low density.Pinus merkusii, which dominates the land, also has a small leaf area, so the canopy density that covers the land tends to be low.

Suggestion
Gunung Halimun Salak National Park can potentially mitigate climate change by increasing carbon stocks.Information on carbon storage in GHSNP still needs to be improved, so research on carbon stock must be carried out at other resorts and other zones in GHSNP.Enrichment planting is required on sites with low species diversity in the GHSNP utilization zone.The species to be planted need to be carefully considered, such as GHSNP endemic species, and not using invasive alien species.

Figure 1 .
Figure 1.Study location in the utilization zone of GHSNP.
Total basal area of the species Total basal area of all species × 100%

Figure 4 .
Figure 4. Distribution of individual trees based on diameter class.

Table 2 .
NDVI values for each density class.

Table 3
shows that the highest number of species was in class 3, with 21 undergrowth and 15 species of woody plants.It shows that the land with the highest vegetation density is more diverse than the other density classes.Understorey species showed the highest number of species compared to woody plant species.The existence of understorey that dominates can be a competitor for tree rejuvenation growth, especially on monoculture land.The lowest number of species was found at the tree level because the GHSNP utilization zone is a homogeneous ecosystem based on the distribution of vegetation types.The Important Value Index (IVI) describes how important a plant species is in a forest community by analyzing species dominance in a community.The species with the highest IVI values in each class are presented in Table4.

Table 7 .
Result of Pearson correlation test.

Table 8 .
Result of the validation test

Table 8
shows that the five prediction models have high accuracy.The validity test results show the model's accuracy, which has a small percentage value[23].