IOP Conference Series: Earth and Environmental Science

PAPER • OPEN ACCESS You may also like

- Predictability of tropical vegetation
Analysis Of Vegetation Index For Ndvi, Evi-2, And greenness using sea surface temperatures
Binyan Yan, Jiafu Mao, Xiaoying Shi et al.
Savi For Mangrove Forest Density Using Google - Comparison of atmospheric correction
models: FLAASH and 6S code and their
Earth Engine In Lembar Bay, Lombok Island impact on vegetation indices (case study:
paddy field in Subang District, West Java)
F Muchsin, D Dirghayu, I Prasasti et al.
To cite this article: Srirahadita Pamungkas 2023 IOP Conf. Ser.: Earth Environ. Sci. 1127 012034
- Mangrove Leaf Area Index Estimation
Using Sentinel 2A Imagery in Teluk Ratai,
Pesawaran Lampung
Anwar Juniansah, Galih Candra Tama,
Kadek Risna Febriani et al.
View the article online for updates and enhancements.

This content was downloaded from IP address 168.151.97.46 on 21/01/2023 at 13:28

Geomatics International Conference 2022 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 1127 (2023) 012034 doi:10.1088/1755-1315/1127/1/012034

Analysis Of Vegetation Index For Ndvi, Evi-2,And Savi For

Mangrove Forest Density Using Google Earth Engine In
Lembar Bay, LombokIsland

Srirahadita Pamungkas
Department of Geomatics Engineering, Sepuluh Nopember Institute of Technology,
Surabaya, 60111, Indonesia
*E-mail: dhita.ayoyi@gmail.com

Abstract. The mangrove forest is one of the most productive ecosystems in the world.
Mangrove forests have benefits such as flood control, groundwater conservancy,
shoreline and storm protection, biodiversity conservation, mitigation, and adaptation to
climate change. Remote sensing for monitoring and mapping natural ecosystems such
as mangrove forests has increased recently. The advantage of using remote sensing data
for mapping mangrove forests lies in the remote sensing imagery that provides a
comprehensive view compared to land-based measurements. The easily-recognizable
appearance of mangrove vegetation in the image is because mangrove vegetation lives
between land and sea transitions; thus, the mangrove vegetation has a darker color
appearance. The vegetation index algorithm is an algorithm that can see the condition
and density of mangrove forests. Therefore, the Google Earth Engine platform can be
utilized. It has a very large remote sensing data set; thus, it can process and discover the
density value of mangrove forests the results of the vegetation index values with the
NDVI, EVI-2, and SAVI methods. The location of this research is in Lembar Bay. The
results of this study indicate that sentinel imagery 2A level 2A can be used to generate
vegetation index data using the NDVI, EVI-2, and SAVI algorithms. Based on the
vegetation index classification with NDVI, EVI-2, and SAVI, it is known that the
dominant density level is a very high density, with an area of 56,66 ha for NDVI, 50,24
ha for EVI-2, and 56,65 ha for SAVI. Then the correlation between NDVI, EVI-2, and
SAVI with water and soil parameters that have the most influence on mangrove
vegetation density is water pH with a correlation coefficient value of NDVI 0,464, EVI-
2 0,469, and SAVI 0,464, showing that water pH and vegetation index have a strong
enough correlation.

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
Geomatics International Conference 2022 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 1127 (2023) 012034 doi:10.1088/1755-1315/1127/1/012034

1. Introduction
The mangrove forest is one of the most productive ecosystems in the world. Mangrove forests have
benefits such as flood control, groundwater conservancy, shoreline and storm protection, biodiversity
conservation, mitigation, and adaptation to climate change [1]. In 2015 the area of mangrove forest in
Lembar Bay was 129.29 Ha and in 2019 there was a decrease, the area of mangrove forest become
120.96 Ha, there was a reduction in mangrove forest due to the activities of Lembar Port, and changes
in land use [2]
Remote sensing has progressed since the second half of the 20th century, mapping and monitoring
of natural ecosystems, including mangrove forests, have been improved in recent decades [3]. One of
the advantages of using remotely sensed data in mangrove research is that it provides a synoptic view
over a large area compared to land-based measurements [4]. Therefore, satellite images with medium
resolution can be accessed easily and free of charge, for example, Sentinel-2A imagery can be used for
mapping and monitoring mangrove forests in large areas [5]. Sentinel 2A image with multispectral
instruments has 13 spectral channels with a spatial resolution of 10 meters. The most widely used image
transformation for mapping mangroves is the vegetation index [6]. Various kinds of vegetation indices
produce different accuracy values for the density of mangrove forests. The NDVI (Normalized
Difference Vegetation Index) vegetation index algorithm is an algorithm that has better results in the
density of mangrove forests [7].
Google Earth Engine (GEE) is a cloud-based platform that makes it easy to access and process very
large geospatial datasets for analysis. The platform allows users to create and run custom algorithms and
fast computations that enable global-scale analysis easily. By using the Google Earth Engine platform,
you can analyze the vegetation index with the NDVI, EVI-2, and SAVI algorithms to determine the
density of mangrove forests using the Sentinel 2 level 2A images.
Among the vegetation indices with NDVI, EVI-2 and SAVI have different value ranges where the
range of mangrove vegetation density values using the NDVI vegetation index has a smaller value range
than EVI-2 and SAVI which have a maximum value range of more than 1 [8].
This study aims to examine the value of the vegetation density index of mangrove forests in Lembar
Bay using Sentinel 2 Level 2A satellite imagery by utilizing the Google Earth Engine platform. And the
results of the classification of the spectral values of the vegetation can be used to make a map of
mangrove vegetation to provide information on the value of the mangrove forest vegetation index in
Sheet Bay.

2. Method
2.1. Research Location
This research was conducted in Lembar Bay, West Lombok Regency, West Nusa Tenggara Province,
the geographical location of Lembar Bay is located between 8° 43' 16.59'' South Latitude to 116° 03'
36.29'' East Longitude and 8° 45' 01.05'' South Latitude to 116° 02' 08.53'' East Longitude. The coastal
villages in the research location are South Lembar Village, Labuan Tereng Village, Eyat Mayang
Village, Cendi Manik Village, Central Sekotong Village, and West Sekotong Village, which can be seen
in Figure 1. below:

2
Geomatics International Conference 2022 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 1127 (2023) 012034 doi:10.1088/1755-1315/1127/1/012034

Figure 1. Map of The Research Location

2.2. Data and Equipment
The data used in this research were sentinel 2A level 2A images for the year 2021, the border checker
data for the administrative area of West Lombok Regency were sourced from the Indonesian Earth Map
with a scale of 1:25,000. Infrared aerial photograph data and sample data of water and soil parameters
whose locations can be seen in Figure 1. Satellite image data were obtained directly from the google
earth engine dataset. The equipment used for water and soil sampling (GPS, plastic bottles, soil drills,
plastic bags, and small shovels) then the water and soil samples will be tested in the laboratory. The
google earth engine software was also used to process satellite images to produce NDVI, EVI-2, and
SAVI which will then be tested with field sample data.
2.3. Image Data Processing
The data used were sentinel 2a Level 2a images. The radiometric correction was not performed on the
satellite image of Sentinel-2 level 2A, because the level 2A product had been radiometrically corrected
and had a Bottom of Atmosphere reflectance value [9]. Then the image was cut with an RBI map
according to the research location which will then be calculated with the vegetation index algorithm
using the NDVI, EVI-2, and SAVI algorithms to get the density of mangrove vegetation. This vegetation
index processing uses data from band 4 (Red) and band 11 (Near Infrared) Sentinel 2A images. The
following is the algorithm for the three vegetation indices:
𝜌𝑁𝐼𝑅 − 𝜌𝑅𝑒𝑑
𝑁𝐷𝑉𝐼 = (1)
𝜌𝑁𝐼𝑅 + 𝜌𝑅𝑒𝑑
𝜌𝑁𝐼𝑅 − 𝜌𝑅𝑒𝑑
𝐸𝑉𝐼2 = 𝐺 (2)
𝜌𝑁𝐼𝑅 + 𝜌𝑅𝑒𝑑
(1 + 𝐿)(𝑁𝐼𝑅 − 𝑅𝑒𝑑)
𝑆𝐴𝑉𝐼 = (3)
𝑁𝐼𝑅 + 𝑅𝑒𝑑 + 𝐿
3
Geomatics International Conference 2022 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 1127 (2023) 012034 doi:10.1088/1755-1315/1127/1/012034

2.4 Field Data Processing

Field data collection was carried out using a drone with an infrared camera. MAPIR Survey3 was carried
out from 19 April 2022 to 21 April 2022. Data collection of water and soil samples was carried out at 8
location points spread over the research location, then to get the data values of water and soil samples,
testing was carried out at the Chemistry Laboratory and Soil Physics Laboratory, University of Mataram.
The following is the distribution of the field data sample points.

3. Results and Discussion

3.1. Vegetation Index Results
In this research, the vegetation index algorithm used is NDVI (Normalized Different Vegetation Index),
EVI-2 (Enhanched Vegetation Index-2), and SAVI (Soil Adjusted Vegetation Index). The following are
the results of the calculation of the NDVI, EVI-2, SAVI vegetation algorithm on the image

Table 1. Calculation Of NDVI, EVI-2, And SAVI Vegetation Index Values on Images
Vegetation Index Value on Sentinel Image 2A level
Vegetation 2A
Index
Minimum Maximum Mean Standard Deviation
NDVI -0,595 0,848 0,024 0,118
EVI-2 -0,702 1,918 0,044 0,236
SAVI -0,892 1,273 0,036 0,177

The resulting vegetation index values have a fairly wide range (between the minimum and maximum
values) and with a standard deviation value that is greater than the average value, even a maximum
value of more than 1.0 is obtained from the EVI-2 and SAVI vegetation indices. The comparison between
EVI 2 and NDVI values yields higher values because EVI 2 has the abilityto capture changes in vegetation
conditions and structures, especially to differentiate between surface greenness and leaf area index for
vegetation with different soil background reflectance. The SAVI vegetation index has a higher value
than NDVI, the SAVI algorithm has included a correction factor forthe influence of soil factors, namely
using the vegetation isoline equation derived from the canopy reflectance approach based on the first-
order photon interaction model between the canopy and the soillayer [10]. Thus, the factors that affect
the value of the vegetation index due to the diversity of vegetationsurface conditions can be reduced.
The next step is to calculate the mangrove area based on its density level. Mangrove density levels
are distinguished based on the range of vegetation index values divided into five classes, namely, very
rarely, rarely, moderate, high density, very high density. Following are the results of mangrove
calculations based on their density level using NDVI, EVI-2, and SAVI classified image data.

4
Geomatics International Conference 2022 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 1127 (2023) 012034 doi:10.1088/1755-1315/1127/1/012034

Table 2. Mangrove Area Based on Density Level

Mangrove Area Based on Density Level (ha)
Index Very
Very High Total Area
Vegetation Rarely Moderate High
Rarely Density (ha)
Density
NDVI 33,300 27,928 35,023 41,594 56,665 194,510
EVI2 39,048 29,394 36,830 39,050 50,242 194,564
SAVI 33,298 27,911 34,982 41,654 56,656 194,501

Graph of Comparison of Mangrove Vegetation Density Based on Density Level

60.000

50.000
Mangrove Forest Density Area (Ha)

40.000

30.000 Series1
NDVI

Series2
EVI-2
20.000
Series3
SAVI

10.000

0.000
0 1 2 3 4 5 6
Mangrove Density Level

Figure 2. Comparison Graph of Mangrove Vegetation Density Based on Density Level

From Table 2 and Figure 2 above, it can be seen that there are differences in mangrove area in the
vegetation index processing results with the NDVI image, EVI 2 image, and SAVI image algorithms.
The results of the mangrove forest density area using the NDVI and SAVI algorithms have area values
that are close to each level of mangrove forest density compared to the results of the EVI-2 algorithm.
Although there are wide differences between each algorithm, it can be seen that the mangrove vegetation
density which has the smallest area is in the rarely class and the widest vegetation density is in the very
high density class. The following is a map of mangrove density using the vegetation index NDVI, EVI-
2 and SAVI.

5
Geomatics International Conference 2022 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 1127 (2023) 012034 doi:10.1088/1755-1315/1127/1/012034

Figure 3. Mangrove Density Map Based on EVI-2 Algorithm

Figure 4. Mangrove Density Map Based Figure 5. Mangrove Density Map Based
on NDVI Algorithm on SAVI Algorithm

6
Geomatics International Conference 2022 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 1127 (2023) 012034 doi:10.1088/1755-1315/1127/1/012034

3.2. Correlation Analysis of NDVI, EVI-2, and SAVI Images with NDVI, EVI-2, and SAVI
Aerial Photographs in the Field
After obtaining the results of processed orthomosaic infrared aerial photographs that have been
calibrated, then calculating the value of the vegetation index algorithm, namely NDVI, EVI-2, and
SAVI. The calculation of the vegetation index is used as validation of the vegetation index value
calculation of NDVI, EVI2, and SAVI on the satellite image of Sentinel 2a Level 2A for the year 2021
that has been processed. The following is the result of the vegetation index values calculation of NDVI,
EVI-2, and SAVI on Infrared aerial photographs:

Table 3. NDVI Vegetation Index Values on Infrared Aerial Images and Photographs
Coordinate WGS 1984 NDVI EVI-2 SAVI
Points UTM 50 S Infrared Infrared Infrared
Image Image
Image Photo Photo Photo
Lat Long
1 398073.0611 9034006.066 0.708 0.847 1.47 0.562 1.06 0.496
2 397325.2604 9030726.859 0.503 0.868 0.933 0.602 0.754 0.269
3 396118.9984 9030783.477 0.549 0.802 1.045 0.602 0.824 0.472
4 395608.4588 9030650.736 0.391 0.881 0.685 0.56 0.586 0.494
5 398411.39 9034382.218 0.258 0.902 0.425 0.452 0.386 0.41
6 396741.9396 9029539.91 0.302 0.661 0.508 0.368 0.453 0.341
7 394005.7593 9032441.065 0.657 0.917 1.326 0.595 0.986 0.521
8 396440.3604 9034766.033 0.47 0.961 0.858 0.454 0.706 0.413

The NDVI, EVI-2, and SAVI values obtained from aerial photographs with the MAPIR Survey3
Infrared camera were then tested for correlation with NDVI, EVI-2, and SAVI Sentinel 2A images to
determine the closeness between the two samples. The following are the correlation coefficients of the
NDVI, EVI-2, and SAVI Infrared camera photographs and the NDVI, EVI-2, and SAVI Sentinel 2A
images.

Table 4. Correlation Coefficient Values Of NDVI, EVI-2, And SAVI Sentinel Image 2A And NDVI,
EVI-2, And SAVI Infrared Camera Photographs

Index
R
Vegetation
NDVI 0,257
EVI-2 0,155
SAVI 0,472

7
Geomatics International Conference 2022 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 1127 (2023) 012034 doi:10.1088/1755-1315/1127/1/012034

From the Table 4 above, it can be seen that the smallest correlation coefficient value is EVI-2, which
is 0.155, where the closeness relationship level is very low positively, for NDVI, which is 0,257, the
closeness relationship level is low positively, while the SAVI value is 0,472, it is strong enough. This
relationship can be seen from the range of coefficient intervals at 0,2 – 0,4 means the level of correlation
between the two variables is low and 0,4 -0,6 which means the level of correlation between the two
variables is strong enough [11].

3.3. Analysis of Water and Soil Parameter Sample Data in the Field
Water and soil parameter data collection in the field was carried out on April 19, 2022 for location
points1,2,3,4,5,6, and 7, then water and soil sample data collection for locations 8, 9 and 10 was carried
out on 20 April 2022. To get the value from the soil sample, a test is carried out at the Physics and Soil
Conservation Laboratory, University of Mataram. To get the value of the water sample, a test was carried
out at the Soil Chemistry Laboratory of the University of Mataram. Tests at the Soil Physics Laboratory
were carried out to determine the type of soil while testing at the Chemistry Laboratory was carried out
to obtain the values of air pH, EC/salinity, soil moisture content and total nitrogen (N).

Table 5. Correlation Coefficient Values Of NDVI, EVI-2, And SAVI Sentinel Image 2A And Sample
Data of Water and Soil Parameter in The Field

R
Index Vegetation
pH EC KL N Sand Dust Clay
NDVI 0,464 0,242 0,086 0,337 0,145 0,29 0,081
EVI-2 0,496 0,235 0,063 0,352 0,136 0,346 0,125
SAVI 0,464 0,243 0,088 0,337 0,145 0,287 0,079

Based on the description of the correlation between mangrove density and water and soil parameters,
it shows that water and soil parameters in the form of water pH have a strong enough and significant
effect on mangrove density, while EC, nitrogen, and dust soil type have a low and significant effect on
mangrove vegetation density, and water and soil parameters in the form of soil moisture content, sand
soil type, and clay type have a very low and significant effect on mangrove density

4. Conclusion
Based on the classification results of the vegetation index using NDVI, EVI-2, and SAVI on Sentinel
2A Level 2A images, it is found that the density level of the dominant mangrove forest area in Lembar
Bay is very good, from the NDVI image results, it is found that the area of mangrove forest with a very
high density level in area of 56,655ha, EVI-2 of 50,442 ha and SAVI of 56,656 ha. The correlation
results between vegetation index of the NDVI and EVI 2 sentinel 2A level 2A image with the vegetation
index of NDVI, EVI-2 infrared photographs can be concluded to have a low correlation positively while
the correlation between SAVI images and SAVI infrared photographs has strong enough correlation
positively. The correlation between NDVI, EVI-2, and SAVI with water and soil parameters that have
the most influence on mangrove vegetation density is water pH with a correlation coefficient of NDVI
0,464, EVI-2 0,469, and SAVI 0,464, indicating that water pH and vegetation index have a strong
enough relationship.

Acknowledgements
The author would like to thank Professor Bangun Muljo Sukojo who supervised and gave technical
recommendations to improve the research.

8
Geomatics International Conference 2022 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 1127 (2023) 012034 doi:10.1088/1755-1315/1127/1/012034

References
[1] M. Brander, L., et al. 2012. "Ecosystem service values for mangroves in Southeast Asia: A meta-
analysis and value transfer application." Ecosystem Services 1(1): 62-69.
[2] Saraswati, N. A. and R. Saraswati 2019. "Pemantauan Mangrove di Teluk Lembar, Lombok Barat
Menggunakan Landsat Tahun 1995 hingga 2019." Seminar Penginderaan Jauh ke-6 Tahun 2019,
404.
[3] Kuenzer, C., et al. (2011). "Remote Sensing of Mangrove Ecosystems: A Review." Remote Sensing.
Vol. 3, No.5, 878-928.
[4] Thakur, S., et al. (2019). "A Review of The Application Of Multispectral Remote Sensing In The
Study Of Mangrove Ecosystems With Special Emphasis On Image Processing Techniques."
Spatial Information Research, Vol.28, No.1, 39-51.
[5] Veettil, B. K., et al. (2018). "Rapidly diminishing mangrove forests in Myanmar (Burma): a review."
Hydrobiologia, Vol.822, No.1, 19-35.
[6] Wicaksono, P., Danoedoro, P., Hartono, & Nehren, U. 2016. Mangrove biomass carbon stock
mapping of the Karimunjawa Islands using multispectral remote sensing. “InternationalJournal
of Remote Sensing, 37(1), 26-52.
[7] Prameswari, A. A. S. R., Hariyanto, T., & Sidik, F. 2015). Analisis indeks vegetasi mangrove
menggunakan citra satelit ALOS AVNIR-2 (Studi kasus: Estuari Perancak, Bali).
Geoid, 11(1), 40-45.
[8] Prameswari, A. A. S. R., Hariyanto, T., & Sidik, F. (2015). Analisis indeks vegetasi mangrove
menggunakan citra satelit ALOS AVNIR-2 (Studi kasus: Estuari Perancak, Bali).
Geoid, 11(1), 40-45.
[9] Gatti, A., & Bertolini, A. 2015. Sentinel-2 products specification document. Rapport technique.
[10] Huete, A. R. 1988. A soil-adjusted vegetation index (SAVI). “Remote sensing of environment,
25(3), 295-309.
[11] Riduwan “Dasar-dasar Statistika, ” Bandung: Alfabeta, 2003