SPACE BORNE TECHNIQUE TO IDENTIFY THE RELATIONSHIP BETWEEN VEGETATION COVER AND URBANIZATION IN THE CITY OF COLOMBO.

Department of Geography, University of Kelaniya, Sri Lanka 123 . ...................................................................................................................... Manuscript Info Abstract ......................... ........................................................................ Manuscript History Received: 22 March 2019 Final Accepted: 24 April 2019 Published: May 2019

Urban vegetation coverage can be considered as one of the indicators that can be used to identify, the process of urbanization. Most of the time with rapid urbanization, vegetation coverage in a city tend to decrease. Hence, it is explicitly visible that the social background, building density and related factors to the urbanization process, have a high relationship with minimizing city vegetation coverage. There are many techniques that can be used to identify the relationship between the process of urbanization and vegetation coverage. Space borne (Remote Sensing (RS) technique) is one of the tools to identify such kind of phenomena. As the main commercial hub in the island, the city of Colombo has expanded as an agglomeration and also population density is increasing rapidly. Massive constructions have also covered major parts of the city, rapidly. In the past years, vegetation coverage minimizes with these processes. Considering all these facts, this study has attempted to identify the relationship between vegetation coverage and urbanization in the city of Colombo with the help of RS techniques. Hence, to identify the relationship between vegetation cover and urbanization, Landsat 8 OLI TIRS (2014) satellite images were used to construct Normalize Difference Vegetation Index (NDVI) and Urban Index (UI) to fulfill the study aim. The results indicated that, there is a significant negative relationship between urbanization process and vegetation coverage in the city.

Background:-
The direct impact of urbanization is rapidly increasing buildup areas. Expansion of urban areas has profound effects on biodiversity and ecosystem functioning at local, regional, and global scales (Zipperer et al.2000). Land use pattern associated with urbanization process leads to modifications of surface microclimatic and hydrological conditions, including the formation of urban heat islands and changes in surface runoff pattern (Hass, 2013).
The vegetation coverage in a city plays a major role in providing an environment for recreation for the inhabitants. These vegetation areas are highly threatened by the transformed urban land use due to increasing pressure on land in most of the cities (Wijesekera & Manawadu, 2009). The citties' environment, sub urban areas, and towns, rely on vegetation to provide ecosystem functions such as air filtering, temperature amelioration, and water storage, filtration and drainage (Bolund & Hunhammar1999). When considering the vegetation coverage in city areas, it has a societal value in defining nature for millions of dwellers living in cities sustaining community health and well-934 being (Ulrich 1984;Kuo & Sullivan 2001;Fuller et al. 2007), as well as frequently contributing to the conservation estate by associate unique biodiversity (McDowell et al.1991;Lawson et al.2008).
According to Verma et al., (2009) the modern technology of RS allows us to collect numerous physical data rather easily, with speed and on repetitive basis and together with Geographic Information Systems (GIS) helps to analyze the data spatially, offering possibilities of generating various options (modeling), thereby optimizing the whole planning process. Also, RS data of various spatial, spectral, and temporal resolutions have been used to characterize land use and land cover change associated with urban growth and application of RS can greatly enhance our knowledge of urban ecological processes (Buyantuyev et al;. Most of the researchers have investigated, the current progresses in RS and GIS having remarkable enhancement about the availability of high-resolution spatial and attribute data for examining the relationship between buildup areas and vegetation structure in a city. The NDVI is effective in predicting photosynthetic activity, because this vegetation index includes both near infrared and red part of the EMR (Govaerts & Verhulst, 2010). Green leaves usually have low reflectance in the blue and red portions of the spectrum because of chlorophyll absorption, with a slightly higher reflectance in the green, so plants appear green to our eyes. Near infrared radiant energy is strongly reflected from the plant surface and the amount of this reflectance is determined by the properties of the leaf tissues: their cellular structure and the air-cell wall-protoplasm-chloroplast interfaces (Kumar and Silva, 1973). So this study used NDVI to extract the vegetation coverage through the Landsat satellite imagery.
Mapping city built-up areas using moderate resolution RS data such as Landsat TM/ETM+ data is complex, because urban areas contain manmade and natural features like vegetation, water body, bare land etc. (Sinha et al., 2016). Landsat 8 OLI TIRS, olso has the same magnificient capacity of capturing city build-up areas as moderate resolution RS data. These urban regions frequently show heterogeneous spectral characteristics and significant spectral confusion between land cover classes and as a result reduce mapping accuracy (do). UI is one of the methods to identify city buildup areas using satellite imageries. RS can be used to obtain a description of building density with a spectral transformation called UI (Kawamura et al., 1997). It is assumed that high pixel value indicates built-up area intensively.
There are many researches have done, using both physical and human aspect to reveal some important matters in urbanization process of the city of Colombo using space born techniques. But the comparison of UI and NDVI in RS is limited to the city of Colombo. Hence considering above facts, this study attempts to identify the relationship between the vegetation cover and the degree of urbanization in the Colombo city area. Using UI and NDVI this has been done accordingly with the help of Landsat 8 OLI TIRS Imagery 2014.

Aim of the study:-
Identify the relationship between vegetation coverage and urbanization in the city of Colombo, using space borne techniques (Special reference to Colombo city Municiple Wards).

Study area
As the main commercial hub in Sri Lanka, the city of Colombo plays a vital role of rapid urbanization in the country. According to the WGS 1984 coordinate system, Colombo city's absolute situation is expanded in between 6 0 46' to 6 0 58' North Latitude and in between nearly 79 0 00' to 79 0 53' East Longitude (Figure 01).
The city, consists of 47 Municipal Wards (Figure 01and Table 01). Some of these Wards have a high population density while others show comparatively low population density. Also the building density and urbanization process also varies within the Wards. Nearing to the Colombo city harbor area, most of the sky scrapers can be seen. Water bodies like Beira Lake provides cooler environment to the surrounding community within the city. There are a few vegetated areas like Viharamahadevi Public Park, some newly constructed small public park lots, are contributing to make a cooler environment within the city. But this vegetation coverage does not provide sufficient safer and cooler background to avoid the stressful environment generated through traffic congestion, buildings, constructions etc. Considering all these facts, this study has selected the city of Colombo as the study area.   (Figure 03).

Constructing Normalize Difference Vegetation Index (NDVI)
NDVI is a calculation of a simple formula consuming two satellite bands. This index is related to identify vegetation coverage and vegetation reflect. Healthy vegetation reflects very well in the near infrared part of the spectrum. Green leaves have a reflectance of 20 percent or less in the 0.5 to 0.7 micron range (green to red) and about 60 percent in the 0.7 to 1.3 micron range (Near Infrared). The visible channel gives some degree of atmospheric correction. The value is then normalized to the range -1<=NDVI<=1 to partially account for differences in illumination and surface slope (Kidwell, 1994). In RS, NDVI was used to identify the vegetation coverage in a certain area. This study produced NDVI with the help of Landsat 8 OLI TIRS data. To generate NDVI, used band 5 (Near Infrared -NI) and band 4 (Visible Red -R) that are available on Landsat 8 OLI TIRS data. Regression analysis is one of the major statistical methods which can be used to show the relationship between two or more considerable variables. Hence to present the relationship between NDVI and UI, least square regression analysis was calculated. NDVI was taken as the independent variables and UI has been chosen as the dependent variable. This regression calculation carried out according to the Municipal Wards of city of Colombo (Figure 01). Apart from pixel distribution calculation, using the function of "Zonal Statistics as a Table" available in ArcGIS 10.1 software link to calculate the regression analysis. This function provides mean values of NDVI and UI according to Municipal Wards in city of Colombo.

Results:-
As mentioned in methodology when considering the generated grid themes for UI, it shows a remarkable relationship with NDVI ( Figure 04). In reality, the highly populated areas like Ginthupitiya, Maradana, Kochchikade, Maligawatta East and West, Grandpass North, Aluthkade East Wards are being showed high UI values. These areas are having less NDVI values (  These pixel values of UI and NDVI in Wards showed a remarkable relationship. The areas having high UI values, was carried minimum NDVI values. Figure 05 shows the pixel distribution of each variables in each Wards.  When, the middle part of the city is compared, these areas carry a high population density. Also high density of buildings can be seen, because of the population concentration. These areas are situated near the Colombo harbor and most of the industrial buildings can be seen there. Hence, the UI values are high in this area and NDVI is low. Figures 06 and 07 show the difference of vegetation coverage in Viharamahadevi Public Park area which is situated in the middle part of the city and Grandpass area nearing to the Colombo Harbor.  (Table 03). To identify the relationship in-between pixel values of NDVI and UI simple regression analysis method was used. In this situation, the study investigated how the urbanization process effects the vegetation coverage.

Ward ID NDVI UI
According to the generated scatter plot diagram shows negative relationship between NDVI and UI ( Figure 08). To prove that, linear squared regression line were generated and plot on the scatter plot. The NDVI and UI regression line parameters show Ŷ = 102.33 -78.448 X (NDVI). When UI is 0, NDVI is -78.448. When increasing NDVI by one unit, UI will decrease by 102.33 units as an average. Adjusted R 2 is 0.6181. It means 61% of the variation of UI can be explained by the NDVI. Liner correlation coefficient is -0.78. It means there is a high negative linear relationship between NDVI and UI.  This study reveals that, space borne techniques are advanced techniques to identify the relationships in environmental studies.