High Resolution Satellite Imagery for finding Greenness Index and Locations for Planting Avenue Trees

Abstract:
This paper presents the use of GIS as a ‘Decision Support System’ for planting avenue trees. Chennai is taken as the study area. Quick Bird imagery having a resolution of 0.61 m (panchromatic) and 2.4 m (multispectral) and LISS-III imagery having a resolution of 23.5 m (multispectral) and 5 m (panchromatic) are used for our study. The green cover of Chennai is found using the ERDAS software. From this, the Greenness Index is computed.

Locations for planting avenue trees are identified based upon the Greenness Index. Further, ground survey is done for those areas to find out the local topography and soil conditions. The ground water potential and water quality for those areas are analyzed. Depending on these factors the type of tree suitable for a particular area is identified.

1. Introduction:
An important step to preserve our ecological environment is to maintain the green cover of the cities. In addition they also provide wildlife habitat enhancement. Trees can play an important role in reducing carbon dioxide levels in the atmosphere by absorbing carbon dioxide and giving off oxygen. They also help us to combat with the problem of global warming. Indiscriminate felling of trees in the past years for urbanization has resulted in decrease of vegetation in many city regions. The distribution of greenery in the cities can be studied and analyzed using high resolution satellite imagery. Our work deals with determination of the present area covered by greenery, the use of high resolution satellite imagery for finding out suitable locations for placing avenue trees and the use of GIS based ‘Decision Support System’ to find suitable type of avenue tree saplings which can be planted in those areas.

2. Study Area:
Our study area is Chennai, which is situated on the north-east end of Tamil Nadu on the coast of Bay of Bengal. It lies between 12° 9' and 13° 9' of the northern latitude and 80° 12' and 80° 19’ of the southern longitude.

It covers an area of about 176 sq km. Chennai's green canopy is largely rain fed, and the rains have been known to be capricious. Chennai's soil is mostly clay, shale and sandstone.

Figure.1 A Satellite Image of Chennai

3. Methodology:

3.1 Data Collection:
The required data were collected from,

Data bank of Institute of Remote Sensing, Anna University.
Forestry Department, Tamil Nadu.
Department of Environment, Tamil Nadu.

3.2 Formula:
We define the formula of Greenness Index as follows:

Greenness Index of an area =                  Total Vegetated Area
                                               –––––––––––––––––––––––––––––––––––
                                                  Geographical boundary defining the area

3.3 Procedure:
The imagery along with the soil map, road network map and ground water potential map are obtained. The Chennai area is first extracted from the imagery by overlaying the Chennai map on it and cropping the study area. The projection system used in the map as well as the imagery should be the same. Otherwise the map projection is changed in accordance with the imagery using the ENVI software. The extracted image is then rectified.

3.4 Greenness Index:
The Greenness Index is defined as the ratio between the total vegetated area and the total geographical area covered. Low Greenness Index values indicate poor green cover that could be the result of climatic changes. Events that can cause low values include moisture shortages and extreme temperatures and biotic interference. High Greenness Index values might reflect ideal growing conditions.

3.5 Determination of Greenness Index:
To determine the Greenness Index, LISS-III (pan-sharpened) imagery of resolution 5 m is used. We have used ERDAS software to find the Greenness Index. The extracted image is first classified. Classification is the process of sorting pixels into a finite number of individual classes, or categories of data, based on their data file values. If a pixel satisfies a certain set of criteria, then the pixel is assigned to the class that corresponds to that criterion. There are two ways to classify pixels into different categories:

Supervised classification.
Unsupervised classification.

We have done unsupervised classification with ten iterations. Image interpretation is used to find the green covered area of the extracted image. This image is compared with the classified image and in turn the vegetated area in the classified image is found.

The area covered by vegetation is found to be 3369.056 hectares.

The total area under study = 17761.456 hectares.

Thus the Greenness Index of Chennai = 0.1897

3.6 Planting Avenue Trees:
QuickBird imagery having the resolution of 0.61 m in the panchromatic image and 2.4 m in the multispectral image is used to establish the locations for planting avenue trees. The imagery is first imported in ArcMap. The space available on either side of the roads is found out for the study area. The following table is created in ArcGIS. It helps us to identify the species of tree suitable for the particular space available.

Table 1

MAJOR TREES:

Scientific Name	Common Name	Height	Width
Betula nigra	River Birch (single stem)	40’-50’	40’-50’
Carpinus betulus	European Hornbeam	40’-60’	30’-40’
Carpinus betulus ‘Fastigiata'	European Hornbeam (upright form)	35’-40’	20’-30’
Celtis occidentalis	Hackberry	40’-50’	40’-50’
Cladasris lutea*	Yellowwood	30’-50’	40’-50’
Fagus grandifolia	American Beech	50'-90'	50'-75'
Fagus sylvatica	European Beech	50’-75’	40’-60’
Ginko biloba	Ginko (male, fruitless)	50’-80’	40’-80’
Gleditsia tricanthos ‘inermis’	Honeylocust, thornless	50’-70’	35’-50’
Gymnocladus dioicus	Kentucky Coffeetree (male, seedless)	60'-75'	40'-50'
Liqudambar styrciflua ‘Rotundiloba'	Sweetgum (fruitless)	65’-75’	40’-50’
Nyssa sylvatica	Blackgum	40’-70’	35’-45’
Platanus x acerifolia	London Planetree	70’-80’	55’-65’
Quercus alba	White Oak	60'-80'	60'-80'
Quercus lyrata	Overcup Oak	45'-55'	45'-55'
Quercus bicolor	Swamp White Oak	60’-80’	50’-80’
Quercus macrocarpa	Bur Oak	70'-80'	70'-85'
Quercus robur	English Oak	70'-80'	75'-85'
Quercus rubra	Northern Red Oak	60’-80’	45’-60’
Quercus phellos	Willow Oak	60’-75’	40’-60’
Sophora japonica	Japanese Pagoda Tree	40’-70’	30’-40’
Taxodium distichum	Bald Cypress	50'-70'	30'-35'
Tilia tomentosa	Silver Linden	50’-60’	50’-60’
Ulmus americana "Valley Forge"	American Elm	60'-80'	30'-50'
Ulmus parvifolia	Lacebark Elm	40’-45’	45’-50’
Zelkova serrata ‘Village Green'	Village Green Zelkova	50’-60’	50’-60’

Table 2

MAJOR TREES:

Scientific Name	Common Name	Height	Width
Acer campestre	Hedge Maple	30’-35’	30’-35’
Acer ginnala	Amur Maple	15’-20’	15’-25
Acer griseum	Paperbark Maple	20’-30’	15’-25’
Amelanchier laevis*	Allegheny Serviceberry	30’-40’	15’-20’
Carpinus caroliniana	American Hornbeam	20’-40’	20’-30’
Cercis canadensis*	“Redbud Texas White”	20’-30’	15’-30’
Cercis Canadensis*	Eastern Redbud	20’-30’	15’-30’
Chinoanthus virginicus*	Fringetree (tree form)	12’-20’	12’-20’
Cornus florida *	White Flowering Dogwood	20’-30’	20’-30’
Cornus florida ‘rubra’*	Pink Flowering Dogwood	20’-30’	20’-30’
Cornus kousa*	Kousa Dogwood	15’-20’	15’-20’
Crataegus crusgalli ‘inermis'*	Cockspur Hawthorn, thornless	25’-30’	25’-35
Craetaegus virdis	Green Hawthorn	20’-35’	20’-35’
Koelruteria paniculata*	Goldenraintree	30’-40’	30’-40’
Malus x*	Flowering Crabapple	20’-25’	15’-20’
Ostria virginiana	Ironwood	25'-40'	20'-30'
Parriotia persica	Persian Parrotia	20'-40'	15'-30'
Prunus x incamp 'Okame'*	Okame Cherry	15-‘25’	15-‘20’
Quercus myrsinifolia	Chinese Evergreen Oak	30’-35’	30’-35’
Styax japonicus*	Japanese Snowbell	20’-30’	15’-25’
Syringia reticulate*	Japanese lilac	20’-25’	15’-20’

* denotes a flowering tree

3.7 Overlay Analysis and Decision Support System:
Superimposing two or more maps registered to a common coordinate system, either digitally or on a transparent material, for the purpose of showing the relationships between features that occupy the same geographic space is referred to as overlay analysis. If along with the road network map, soil map water quality map and ground water potential map are overlaid then the suitable tree for that particular space, water quality and soil conditions can be identified. For this, overlay analysis is done using ArcMap software. Thus overlay analysis acts as a ‘Decision Support System’ and it helps the user to find out the particular type of tree which satisfies the local conditions.

Figure.2 Overlay Analysis Using ArcMap

4. Results and Discussion:

The Greenness Index for Chennai is found to be 0.1897.
It is established that only 18.97% of Chennai is covered by greenery as against the ideal 33%.
The greenness index can be improved by planting avenue trees.
ARCGIS is used to create a ‘Decision Support System’ that helps the user to take decisions about suitable locations for planting avenue trees.
By comparing the Greenness Index for various years the depletion of avenue trees for different areas can be determined.
The Greenness Index can be monitored every year to out the efficiency of the planting project.
Calculation of Greenness Index for different zones of the city will show the distribution of greenery which will help the administrator to take corrective actions.

5. Conclusion:

From the results mentioned above it is clearly seen that Remote Sensing serves as an important tool in monitoring the green cover of city areas due to temporal data provided by the satellite imagery.
The digital classification methods are based on spectral signature and its accuracy depends upon the quality and quantity of the sample area.
This method saves time and effort for data collection
Ground visits are necessary for verification of data interpreted from satellite imagery.