Response of land surface temperature with the changes of coastal build-up and vegetation index in the mangrove ecosystem of Chattogram coast, Bangladesh

Mangrove vegetation plays a vital role in habitat and nursing ground for different organisms and prevents coastal erosion caused by wave and tide action. In recent years the mangrove vegetation in Chattogram coast, Bangladesh, has been interrupted by other infrastructural development, which has a destructing effect on the surrounding environment. Land surface temperature analysis of an area helps learn about different environmental conditions, weather, and climate. It is also essential to monitor the rising temperature and global warming, the biggest threat to humanity. NDBI and NDVI are the efficient process for monitoring vegetation and build up areas of a geographical location. This study focused on those analyses to understand the importance of mangrove vegetation in the Salimpur area and surrounding coastal areas of Chattogram by studying the relationship between NDVI and NDBI, NDVI and LST, NDBI, and LST. The outcome indicates that a higher vegetation index results in lower land surface temperature during different periods, negatively correlated. This study also found a strong positive correlation between buildup index (NDBI) and land surface temperature (LST), which means Land Surface temperature was found higher in Buildup areas. The vegetation areas are greatly affected by the buildup areas. The correlation between buildup areas and vegetation areas was strongly negative, which means an increase of NDBI decreases NDVI, and a decrease of NDBI increases NDVI.


Introduction
The temperature between the land surface and the earth's atmosphere is known as Land Surface Temperature (LST) [1]. It is frequently employed in all physical parameters, such as balancing the water and energy resources at the interface between the earth's surface and the atmosphere [2]. LST has broad application for assessing vegetation monitoring, soil moisture, evapotranspiration, hydrological cycle, thermal inertia, vegetation water stress, and urban climate [3][4][5]. It is also used to assess the environment and climate change by observing long-term data [6]. LST is also very effective for obtaining latent heat flux [7]. It is necessary to observe the temperature of various land use/cover and is often used for predicting and checking crop yields [8]. LST has changed in time with vegetation changes, soil composition and topography, and land use land cover [9,10]. Generally, the land surface temperature is measured with remotely sensed satellite data as it is impractical to obtain in-situ data regionally or globally. The remote sensing technique is applied to observe spatiotemporal data such as land cover change and basic physical properties [11]. The surface temperature of different land cover being obtained from satellite-based TIR sensors [12]. LST can also be used to observe forest areas, urban areas, desertification, etc., as it is sensitive to soil moisture and vegetation. Multispectral remotely sensed satellite data provide information for analyzing urbanization and desertification effectively [11]. Vegetation cover has a significant influence on land surface temperature. Vegetation regulates the surface thermal activity through evapotranspiration [13][14][15], a process by which heat from the air results in the evaporation of water [16]. The most common vegetation index is the Normalized Difference Vegetation Index (NDVI) which is used for assessing the vegetation condition by using the photosynthetic output of a pixel obtained from satellite data [17]. Unlike vegetation cover, the land surface temperature gets influenced by buildup activity on the earth surface. Urbanization mainly occurs when vegetated surfaces are converted into impermeable built-ups, which refers to the conversion of natural surfaces with different artificial settlements such as residential and industrial infrastructures, roads, bridges, and impervious surfaces [18][19][20]. The transformation affects the humidity in the air, which is influential in landing surface temperature change [21].
In remote sensing-based built-up area assessment, normalized Difference Built-up Index (NDBI) is used to determine the built-up areas. In an urban area, the higher NDBI value refers to the urban areas or built-up areas, and the lower value refers to the vegetated areas [22]. Besides higher NDVI value indicates dense vegetation in a vegetated area. Both NDBI and NDVI significantly influence LST [23]; correlation among these indices indicates how urbanization/built-up and vegetation affect the earth's surface thermal activity and relevant environmental phenomenon. So, it is essential to know the correlation information and how these NDBI, NDVI, and LST influence each other. This study aims to assess the correlation among the NDBI, NDVI, and LST at a coastal region of Salimpur mangrove forest in Chattogram, Bangladesh, which has an artificial mangrove forest with 400 acres. In recent years this area has lost its forest due to urbanization and many natural and artificial activities. The current study demonstrates the relationship between NDBI and NDVI, NDBI and LST, NDVI and LST at different years in the Salimpur region. This research and relevant information will open the window for further study on the environmental impact of urbanization, vegetation, and thermal activity of the earth's surface.

2.1.Study area
The investigation was carried out from November 2016 to November 2020 at the Salimpur mangrove area. The geographical position of the Salimpur mangrove area was latitude 22˚15" N and longitude 91˚49" E and about 15km off from Chittagong Port City ( Figure 1). The study area is about 7.62 sq. km. It is an artificially planted mangrove area situated mainly in the Salimpur union's northern part of the city. Landsat images were further rectified to a standard Universal Transverse Mercator coordinate system using ArcGIS Pro. For further analysis, the satellite images provided by Landsat also undergo atmospheric and geometric correction, applied by images processing in ENVI to improve the quality ( Table 1).

2.3.NDBI Analysis
Compared to other surface features, built-up lands have higher reflectance in the MIR wavelength range (1.55 ~ 1.75μm) than in the NIR wavelength range (0.76 ~ 0.90μm). NDBI is helpful to map urban built-up areas, which is expressed as follows Where NIR is near-infrared reflectance such as ETM+ band 4; MIR is middle infrared reflectance which is ETM+ band 5. NDBI values range from -1 to 1. The greater the NDBI is, the higher the proportion of build-ups [22].
Where NIR and RED refer to the Near Infrared and Red spectral reflectance value, the value of NDVI ranges from +1.0 to -1.0, and the area with a value of NDVI less than -1 or more excellent than +1.0 is considered a No Data zone.

2.5.LST Analysis
The thermal band is used to convert the raw value into the black body temperature in Degree Celsius using ArcGIS Pro software. The OLI thermal infrared band 10 (10. 6-11.19μm) was utilized to derive the LST. The first step is to convert the DN (Digital Number) values of band 10 to atsensor spectral radiance using the following equation [25], After that, the conversion of spectral radiance to temperature in kelvin [18] is-

2.6.Correlation Analysis
Correlation is a bivariate analysis that measures the strength of association between two variables and the direction of the relationship. In terms of the strength of the relationship, the value of the correlation coefficient varies between +1 and -1. A value of ± 1 indicates a perfect degree of association between the two variables. As the correlation coefficient value goes towards 0, the relationship between the two variables will be weaker.
The coefficient sign indicates the direction of the relationship; a '+' sign indicates a positive relationship, and a '-'sign indicates a negative relationship. Usually, in statistics, we measure four types of correlations: Pearson correlation, Kendall rank correlation, Spearman correlation, and the Point-Biserial correlation [26].
Pearson r correlation is the most widely used correlation statistic to measure the relationship between linearly related variables. In this study, we will use the Pearson r correlation method.
Pearson r correlation is used to measure the degree of relationship between the two [27]. The following formula is used to calculate the Pearson r correlation, Where,

Results
The NDBI values in the study area ranges from -0.390 to 0.126. The lowest build-up index was observed in the year 2019 and the highest value was found in the years 2020. Table 3 presents the statistical analysis of the NDBI of Salimpur mangrove vegetation for the years 2016 to 2020. towards the coast in recent years [28].  The NDVI values represented the healthiness of vegetation from the years of 2016 to 2020 at the Salimpur mangrove ecosystem ( Table 4). According to Table 5, the land surface temperature was almost constant in the study periods ranging from 28.75°C to 30.95°C. The mean values of LST showed almost same throughout these years.   and Ethiopia [30][31][32]. The strength and stability of the relationship can be related to the data collection period as all the data were collected mainly from the post-monsoon era [30]. The relationship between vegetation index (NDVI) and surface temperature (LST) of each study year was negative and moderate. This relationship is significant and relevant in the post-monsoon and winter season, supporting the findings of this study [33]. This positive correlation coefficient is also related to evapotranspiration with threshold temperature [34], as the average temperature was higher in these two years. The relationship between NDVI and LST values found in this study is much lower and poor, which can be linked with the urban settlements in and around the study area controlling the land surface temperature variation [35]. A strong negative relationship was found between NDBI and NDVI, which is significant to study the expansion and transformation of builtup areas [11] and plantations and the UHI effect in this vital mangrove ecosystem. Thus, this study revealed that NDBI than NDVI highly influences LST. around 0.813 to 0.895. The current study also showed that the influence of NDBI on LST was more potent than the influence of NDVI on LST. This study suggested for future research that the calculated value of NDBI, NDVI, and LST and their correlation may lead to further research to assess the environmental impacts and climate change assessment.