See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/353337686

Spatio-Temporal Variation of Vegetation and Urban Sprawl Using Remote

Sensing and GIS: A Case Study of Cuttack City, Odisha, India

Article · July 2021

CITATIONS READS

7 757

4 authors, including:

Prasanta Patra Duryadhan Behera

Sambalpur University Sambalpur University
15 PUBLICATIONS 20 CITATIONS 24 PUBLICATIONS 186 CITATIONS

SEE PROFILE SEE PROFILE

Shreerup Goswami
Utkal University
198 PUBLICATIONS 1,750 CITATIONS

SEE PROFILE

All content following this page was uploaded by Prasanta Patra on 20 July 2021.

The user has requested enhancement of the downloaded file.

 Vol. 6, No.2, July, 2021
www.gondwanags.org.in pp. 213-219

Spatio-Temporal Variation of Vegetation and Urban Sprawl Using

Remote Sensing and GIS: A Case Study of Cuttack City, Odisha, India
P.K. Patra, Duryadhan Behera*, S.P. Naik and Shreerup Goswami

Department of Earth Sciences, Sambalpur University, Burla-768019 (OR), India

(*Corresponding author, Email: dbeherasbp@gmail.com)

Abstract

The present study has been conducted to detect the vegetation changes and urban sprawl of Cuttack city, one of the ancient towns of
Odisha, India. Morden and integrated techniques of remote sensing and GIS approaches have been deployed to assess the land use/ land cover
dynamics of the city over 13 years. The proposed study has been undertaken by the help of Landsat -5 TM (2005) and Landsat-8 (OLI/ETM)
(2018), as these are freely available on USGS earth explorer (https://earthexplorer.usgs.gov) Two different periods of Landsat satellite imagery
data i.e., Landsat -5 TM (2005) and Landsat-8 (OLI/ETM) (2018) were examined. The land use/ land cover dynamics of the city over the
duration from 2005 to 2018 was enumerated. Landsat satellite imagery was analysed through advance and sophisticated GIS software like Arc
GIS 10.1 and ERDAS IMAGINE 2013.Using Maximum Likelihood Classification Technique images were classified to find the land use/land
cover for both the study year. The result reveals that built-up area increase by 49.52%. we use DNDVI method to map vegetation change over the
period from 2005 to 2018, DNDVI value between -0.2899 to -0.0145 represent loss of vegetation and threshold value 0.0371 and above shows
vegetation gain.

Keywords: Remote Sensing; GIS; NDVI; Maximum Likelihood Classification, Cuttak City, Odisha

Introduction earth's surface during a specific period of observation (Seto,

Urban sprawling is a recent global trend in the
developing country, leading to alteration of land use/ land
cover of a region (Belal and Moghanm, 2011; Rawat et al.,
2013). The unprecedented population explosion, rapid
industrialisation, unplanned urbanisation, fast movement of
rural people towards urban areas for the search of better job
and opportunity, education, healthcare and lifestyle are chief
factors of urban sprawl (Zhang, 2001; Bekele, 2005; Godfrey
and Julien, 2005; Cohen, 2006; Peng et al., 2011; Iqbal and
Khan, 2014; McGranahan and Satterthwaite, 2014). Urban
sprawl has an immediate impact on the ecology of a region and
modifies the land use and land cover differently. In the
terrestrial ecosystem, Land Use (LU) and Land Cover (LC) are
two basic units having close association with both natural
processes and anthropogenic activities (Jansen and di
Gregorio, 2002; Bender et al., 2005; Mendoza et al., 2010,
2011; Iqbal and Khan, 2014). The land cover often denotes to
vegetation cover, natural earth covers viz. asphalt, trees, dense
forest, grasslands, ploughed land and water bodies, etc. on the
2002; Fisher et al., 2005). Land cover is different from land-
use though they are very often used interchangeably (Fisher et
al., 2005). Land use refers to how the land is being utilized for
socioeconomic purposes, which involves full control of
human beings over land for various purposes (Lambin and
Meyfroidt, 2010; Briassoulis, 2019). Land use and land cover
change (LULC) is a rising topic to monitor the environment,
planning, and management and understanding of many social,
economic and environmental problems (Pelorosso et al., 2008;
Xiao et al., 2006; Gurav et al., 2019). This LULC change
represents how the land is being modified physically and
biologically to obtain food and other essentials over a period of
time.
Monitoring of LULC changes by conventional methods
involves enormous money, human resources, and time.
Satellite Remote Sensing (SRS) has many advantages as this
method is cost-effective, time-saving and provides large scale
data on LULC changes with information about their
geographic distribution (Yuan et al., 2005). The integrated
approach of Remote Sensing (RS) and Geographic
Information Systems (GIS) have proved to be an effective
JGSR, Vol. 6, No.2, July 2021 Spatio-Temporal Variation of Vegetation and Urban Sprawl using Remote Sensing and GIS
Fig. 1. Location Map of the Study Area
technique for assessing the spatiotemporal dynamics of LULC
(Hathout, 2002; Herold et al., 2003; Lambin et al., 2003; Serra
et al., 2008; Iqbal and Khan, 2014). Gathering of information
and updating of LULC maps are essential for management of
natural resources and environmental changes as a changed
dynamics provide critical inputs for planning and management
(Mei and Qing, 1999; El-Kawy et al., 2010; Kamp et al., 2008;
Owen et al., 2008). Since ecosystems are ever-changing,
LULC changes are therefore categorised in two ways, namely
“land cover modifications” and “land cover conversions”
(Coppin et al., 2004, Henebry and Goodin, 2006). Coppin et
al. (2004) defined the land cover conversions as “the complete
replacement of one cover type by the other” and land cover
modifications as “the subtle changes that affect the character
of land cover without changing its overall classification."
Land cover conversions are the results of urban expansion,
wildfires, agricultural expansion whereas land cover
modifications are mainly results of climatic variability,
pollution, intra-annual cycles (Ahl et al., 2006; Lunetta et al.,
2006, Telesca and Lasaponara, 2006; Ghaznavi et al., 2018).
As a consequence, these two types of land cover changes have
implications on various aspects of studies.
Our environment is getting deteriorated due to
population explosion, rapid depletion of natural resources,
pollution, unplanned land use, and various other reasons
(Satapathy et al., 2007). Hence, it is very much essential to
derive information on the change in land use and land cover for
different kinds of studies. This information plays a vital role in
various local and regional regulatory and governing bodies for
macro-level planning. Various planning and management
activities are incompetent due to a lack of information on rates
of land cover/land-use change. The land-cover changes occur
p 214
naturally progressively and gradually, though it may be rapid
and abrupt due to anthropogenic activities.
Location of the Study Area
Odisha is located in the eastern part of the Indian
subcontinent bordered by West Bengal, Chhatisgarh,
Andhrapradesh and Jharkhand. Geographically, it is located in
the tropical zone between latitude 17° 47′ N and 22° 34′ N and
longitude 81° 22′ E and 87° 29′ E. The Cuttack city is selected
as the study area as it is one of the ancient and fast-growing
towns of Odisha, India (Fig. 1). The city is the district
headquarters and well connected with all parts of the country
through SE railway and NH-5 and also with overseas as Biju
Pattanaik International airport, Bhubaneswar is at 35 km away.
The city carries a lot of historical importance and value in
Odisha's history. Cuttack is also called as millennium city and
silver city due to its 1000 years of remarkable history and
fabulous silver filigree work. The most important part of the
city is that it is centrally located between the Kathajodi and the
Mahanadi River.
Methodology
The 2005 Landsat-5 data and 2018 Landsat-8 data were
downloaded via https://earthexplorer.usgs.gov by putting the
required information (desired location and time). For the
processing of these data, ERDAS IMAGINE 2013 and Arc
GIS 10.1 were used. The detailed technical specification is
given below (Table 1).
The method on which the study was made is shown with
a flow chart (Fig. 2). Satellite data from both the year is
215 p P.K. Patra et al. JGSR, Vol. 6, No.2, July 2021

Table 1: Satellite data used for LULC study.

Name of the Sensor Date of Resolution
satellite used acquisition (m)
Landsat-5 MSS/TM 21/01/2005 30
Landsat-8 OLI/TIRS 07/03/2018 30

Normalized Difference vegetation index (NDVI)

Vegetation indices allow us to delineate the distribution

Fig. 2. Flow chart showing the different stages of data analysis.
processed using image processing software ERDAS imagine
2013; the different bands of the data are stacked and cropped
by the study area boundary. NDVI (Normalized Difference in
Vegetation Index) was found using the NIR and RED band of
Landsat satellite data for the year 2005 (Fig. 3.) and 2018 (Fig.
4) in ERDAS IMAGINE version 2013 and maximum likely
hood classification is done using Arc GIS 10.1. In order to find
the change in urban growth, both the supervised raster maps
from 2005 and 2018 are converted to vector maps using Arc
GIS 10.1 software and urban area extracted from these maps
was analysed to detect the urban sprawl. This method helped to
delineate the urban sprawl growth and the direction in which
the sprawling growth is progressive. To find the change in
vegetation DNDVI method is used.
DNDVI = NDVI (t2) – NDVI (t1)
Here t1is 21 may 2005 and t2 is 07 march 2018.
DNDVI value between -0.2899 to -0.0145 shows loss of
vegetation and threshold value 0.0371 and above shows gain
in vegetation.
of vegetation and soil based on the characteristic reflectance
patterns of green vegetation (Gandhi et al., 2015). NDVI is a
powerful index to detect the vegetation change over the year.
The sudden increage of near-infrared reflection of vegetation
from red reflection is very useful to study and detect the
vegetation status of an area.
NDVI = (NIR-RED)/(NIR-RED)
Where NIR-near-infrared (750-1300 nm)
RED-visible red band (600-700 nm)
NDVI value ranges from +1 to -1. The very low value of
NDVI<=0.1 corresponds to snow, sandy area, or barren land,
while high NDVI>=0.6 to 0.8 corresponds to dense vegetation
and for bare soil area NDVI tends toward 0 and for water
NDVI is negative.
DNDVI= {(NIR-R)/(NIR+R)}(t2) - {(NIR-R)/(NIR+R)}(t1)
Maximum Likelihood Classification
Stacked Landsat 5 and Landsat 8 satellite images are
classified using image classification tool in Arc GIS 10.1
environment. A signature file is created by selecting the
training sites for the built-up area, barren land, vegetation,
water etc. based on ground truth and expertise over image
interpretation then, maximum likelihood classification is
done.
Accuracy Assessment
To assess the accuracy of the present classification, the
classified image of 2018 is compared with verification in the
ground. A total of fifty numbers of ground-truthing (verifying)

Fig.3. Vegetation covers in the area during the year 2005 Fig.4. Vegetation covers in the area during the year 2018
JGSR, Vol. 6, No.2, July 2021 Spatio-Temporal Variation of Vegetation and Urban Sprawl using Remote Sensing and GIS p 216

points were collected using GPS from different locations Table 3: Land use/land cover change classes
within the study area. Then these points were superimposed Landuse Change Percentage change
Year
over the classified image on Arc GIS 10.1, and by using classes in area in area
various tools like "extract value to point" and "pivot table," a
2005 2018 2005-2018 2005-2018
confusion matrix is generated. Finally, User's Accuracy,
Producer's accuracy and Overall accuracy is calculated (Table Builtup land 19.47 29.11 -9.64 -49.52
2). For accuracy assessment of classified image of 2005, we Vegetation 20.07 13.01 7.06 35.17
Barren land 5.76 3.02 2.74 47.63
used the historical image of study area for the year 2005 from
Water 0.41 0.57 -0.16 -38.72
google earth and used it for finding accuracy.
N.B-Area in Sq.km
Formula for overall accuracy is given by:
OverallAccuracy% = (Total number of true predictions / and subsequently vegetation and barren land reduce by
total number of samples) X 100 35.17% and 47.63 % from 2005 to 2018 (Fig. 11) (Table 3
It has been revealed that the current estimation of LULC and 4). To show how the LULC change from one category to
dynamic using the latest technology is about 90.56 % and other from 2005 to 2018 (Fig. 12) here created a cross
93.25 % accurate for the year 2005 and 2018, respectively. tabulation matrix (Table 5).
DNDVI value is useful in finding vegetation loss and
Kappa coefficient is generated from statistical test to vegetation gain area during the study period. Threshold value
evaluate the accuracy of a classification. 0.0371 and above shows vegetation gain and area between
Kappa coefficient = Total Accuracy-Random Accuracy / value -0.2899 to -0.0145 shows loss in vegetation (Fig. 13).
1-RandomAccuracy
Conclusions
Our classification result found Kappa value 0.825,
which demonstrate that our classification comes under very The present study reveals that the city kept its
good category. importance over the years as a result of which urban growth is
evident. The spatiotemporal monitoring of vegetation and
Results and Discussion urban sprawl shows unprecedented growth in an urban area,
which causes loss of vegetation coverage. Urban area, which
Cuttack city, one of the most growing cities of Odisha was 42.59 % in the year 2005 that has been sprawled to 63.68%
and also a business hub; its urban landscape has grown over the in the year 2018, indicates the pace of urbanisation over the
year as the present study revealed the same. On the other hand, study period. The vegetation coverage has also depleted from
the maximum likelihood classification of the study area for the 20.07% in the year 2005 to 13.01% in the year 2018. It is a
years 2005 (Fig. 5) and 2018 (Fig. 6) also shows a drastic matter of grave concern to act upon the impact of urbanisation
2
increase in the built-up area from 19.47 km in 2005 (Fig.7) to
2
29.11 km in 2018 (Fig. 8). The classified map of 2005 (Fig. 9)
Table 4: Land use/land cover change classes
and 2018 (Fig. 10) justified our finding where the vegetation
coverage and barren land have been reduced. The statistics Study Year Percentage of area
LULC
2005 2018 2005 2018
shows 49.52% increase in built up area during study period
Builtup land 19.47 29.11 42.59 63.68
Vegetation 20.07 13.01 43.90 96.04
Table 2: Showing the accuracy of LULC in the study area. Barren land 5.76 3.02 12.60 6.60
Water 0.41 0.57 0.91 1.26
Results of Accuracy Assessment
N.B-Area in Sq.km
Sr. Accuracy
Land cover 2005 2018
No. Assessment
1 Producer Accuracy Built up area 92% 94% Table 5: Cross Tabulation Matrix
Vegetation 87% 82% Land use 2018
Water 89% 94%
Barren land 91% 93% Barren Builtup Vegetation Water Total
land land
2 User Accuracy Built up area 88% 92%
Vegetation 92% 94% Barren land 1.78 2.06 1.88 0.03 5.76
Land use Builtup land 0.57 16.67 2.01 0.22 19.47
Water 90% 92%
2005 Vegetation 0.66 10.15 9.04 0.22 20.07
Barren land 89% 91%
Water 0.01 0.23 0.08 0.09 0.41
3 Overall Accuracy 90.56% 93.25%
Total 3.02 29.11 13.01 0.57 45.71
217 p P.K. Patra et al. JGSR, Vol. 6, No.2, July 2021

Fig.5. Map showing urban area in the year 2005 Fig.6. Map showing urban area in the year 2018

Fig. 7. Land use land cover pattern during the year 2005 Fig.8. Land use land cover pattern during the year 2018

Fig.9. Map showing the land use land covers for the year 2005 Fig.10. Map showing the land use land covers for the year 2018

Fig.11. Showing different Land use land covers area for the year 2005 and 2018

Fig. 12. Showing extent of urban area during year 2005 and 2018 Fig. 13. Map showing vegetation loss and gain area during 2005 to 2018
JGSR, Vol. 6, No.2, July 2021 Spatio-Temporal Variation of Vegetation and Urban Sprawl using Remote Sensing and GIS
on the ecosystem. The 7% decrease in vegetation cover over
13 years is alarming. Therefore, the town planners and
administrators are required to implement suitable measures to
increase the vegetation cover for improvement of the
microclimatic conditions of the city.
Authors' Contributions
P. K. Patra: Investigation, Conceptualization, Formal
Analysis. Duryadhan Behera: Data Collection, Investigation,
Conceptualization, Formal Analysis, Writing-Original Draft,
Reference
Ahl, D.E., Gower, S.T., Burrows, S.N., Shabanov, N.V., Myneni, R.B. and
Knyazikhin, Y. (2006). Monitoring spring canopy phenology of a
deciduous broadleaf forest using MODIS. Rem. Sens.
Environment., v.104, pp.88-95.
Bekele, H. (2005). Urbanization and urban sprawl. Royal Institute of
Technology: Stockholm, Sweden, 248p.
Belal, A.A. and Moghanm, F.S. (2011). Detecting urban growth using
remote sensing and GIS techniques in Al Gharbiya governorate,
Egypt.The Egyptian Jour. Rem. Sens. Space Sci., v. 14(2), pp. 73-
79.
Bender, O., Boehmer, H.J., Jens, D. and Schumacher, K.P. (2005). Using
GIS to analyse long-term cultural landscape change in Southern
Germany. Landsc. Urban Plan., v. 70, pp. 111-125.
Briassoulis, H. (2019). Analysis of land use change: theoretical and
modelling approaches. v.294, pp.1-65.
Cohen, B. (2006). Urbanization in developing countries: Current trends,
future projections, and key challenges for sustainability. Technol.
Soc., v. 28(1-2), pp. 63-80.
Coppin, P., Jonckheere, I., Nackaerts, K. and Muys, B. (2004). A review:
Digital change detection methods in ecosystem monitoring: a
review. Intl. Jour. Rem. Sens., v. 25, pp. 1565-1596.
El-Kawy, O.R.A., Rod, J.K., Ismail, H.A. and Suliman, A.S., (2010).
Land use and land cover change detection in the Western Nile Delta
of Egypt using remote sensing data. Appl. Geogr., v. 31, pp. 483-
494.
Fisher, P.F., Comber, A.J. and Wadsworth, R. (2005). Land use and land
cover: contradiction or complement. Re-presenting GIS, pp.85-98.
Gandhi, G.M., Parthiban, S., Thummalu, N. and Christy, A. (2015).
NDVI: Vegetation change detection using remote sensing and GIS-
A case study of Vellore District. Proced. Comput. Sci., v. 57, p.
1199-1210.
Ghaznavi, A. A., Ahmad, A. H. M., Alam, M. M. and Quasim, M. A.
(2018). Heavy mineral study of Dhosa Sandstone, Kachchh Basin,
Western India: Implication for provenance and tectonic setting.
Jour. Geosci. Res., v. 3(1), pp.1-12.
Godfrey, R. and Julien, M. (2005). Urbanisation and health. Clinical
Med.., v. 5(2), p.137.
Gurav, C., Babar, M. and Khan, I. (2019). Identification of Groundwater
p 218
Supervision, Reviewing and Editing. Shreerup Goswami:
Writing-original draft. S. P. Naik: Data collection, Reviewing
and Editing.
Acknowledgements
The authors would like to extend heartfelt thanks to both
the anonymous reviewers and editor for spending their
valuable time to review this manuscript. The authors also wish
to thank the authorities of the Sambalpur University for
providing the platform to carry out research.
Potential Zones in Manar River Sub-Basin, Maharashtra using
Remote Sensing and GIS Techniques. Jour. Geosci. Res., v. 4(1),
pp.31-38.
Hathout, S. (2002). The use of GIS for monitoring and predicting urban
growth in East and West St Paul, Winnipeg, Manitoba, Canada.
Jour. Environ. Manag., v. 66, pp. 229-238.
Henebry, G.M. and Goodin, G.D. (2006). Landscape Trajectory Analysis:
Toward Spatio-temporal Models of Bio-geophysical Fields for
Ecological Forecasting.Accessed on December 08, 2006.
Herold, M., Goldstein, N.C. and Clarke, K.C. (2003). The spatio temporal
form of urban growth: measurement, analysis and modelling. Rem.
Sens. Environ., v. 86, pp.286-302.
Iqbal, M.F. and Khan, I.A. (2014). Spatiotemporal land use land cover
change analysis and erosion risk mapping of Azad Jammu and
Kashmir, Pakistan. The Egyptian Jour. Rem. Sens. Space Sci., v.
17(2), pp. 209-229.
Jansen, L.J.M. and di Gregorio, A. (2002). Parametric land cover and land
use classification as tools for environmental change detection.
Agric. Ecosyst. Environ., v. 91, pp. 89-100.
Kamp, U., Growley, B.J., Khattak, G.A. and Owen, L.A. (2008). GIS
based landslide susceptibility mapping for the 2005 Kashmir
earthquake region. Geomorphology, v.101, pp.631-642.
Lambin, E.F. and Meyfroidt, P. (2010). Land use transitions: Socio-
ecological feedback versus socio-economic change. Land Use Pol.,
v. 27(2), pp. 108-118.
Lambin, E.F., Geist, H.J. and Lepers, E. (2003). Dynamics of land-use
and land-cover change in tropical regions. Ann. Rev. Environ.
Resour., v. 28, pp. 205-241.
Lunetta, R.S., Knight, J.F., Ediriwickrema, J., Lyon, J.G. and Worthy,
L.D. (2006). Land-cover change detection using multi-temporal
MODIS NDVI data. Rem. Sens. Environment., v. 105, pp. 142-154.
McGranahan, G. and Satterthwaite, D. (2014). Urbanisation concepts and
trends. London: Iied, 28p.
Mei, X. and Qing, R. (1999). Change detection based on remote sensing
information model and its applications on coastal line of Yellow
River Delta. Earth Observation Centre, NASDA, China.
Mendoza, M., Granados, E., Geneletti, D., Perez-Salicrup, D. and
Salinas, V. (2011). Analysing land cover and land use change
 219 p P.K. Patra et al.
processes at watershed level: a multi temporal study in the Lake
Cuitzeo Watershed, Mexico (1975-2003). Appl. Geogr., v. 31, pp.
237-250.
Mendoza, M.E., Granados, E.L., Geneletti, D., Diego, R., Perez-Salicrup,
D.R. and Salinas, V. (2010). Analysing land cover and land use
change processes at watershed level: a multi temporal study in the
Lake Cuitzeo Watershed, Mexico (1975-2003). Appl. Geogr., v. 31,
pp. 237-250.
Owen, L.A., Kamp, U., Khattak, G.A., Harp, E.L., Keefere, D.K. and
Bauerf, M.A. (2008). Landslides triggered by the 8 October
2005Kashmir earthquake. Geomorphology, v. 94 (1-2), pp. 1-9.
Pelorosso, R., Leone, A. and Boccia, L. (2008). Land cover and land use
change in the Italian centralApennines: a comparison of assessment
methods.Appl. Geogr., v. 29 (1), pp. 35-48.
Peng, X., Chen, X. and Cheng, Y. (2011). Urbanization and its
consequences. Paris, France: Eolss Publishers, v.2, pp.1-16.
Rawat, J.S., Biswas, V. and Kumar, M. (2013). Changes in land use/cover
using geospatial techniques: A case study of Ramnagar town area,
district Nainital, Uttarakhand, India. The Egyptian Jour. Rem. Sens.
Space Sci., v. 16(1), pp.111-117.
Satapathy, D.R., Krupadam, R.J., Kumar, L.P. and Wate, S.R. (2007). The
application of satellite data for the quantification of mangrove loss
and coastal management in the Godavari estuary, East Coast of
(Received: 31 March 2021; Revised Form Accepted: 20 May 2021)
View publication stats
JGSR, Vol. 6, No.2, July 2021
India. Environment. Monitor.Assess., v.134 (1-3), pp. 453-469.
Serra, P., Pons, X. and Saurı, D. (2008). Land-cover and land-use change
in a Mediterranean landscape: a spatial analysis of driving forces
integrating biophysical and human factors. Appl. Geogr., v.28,
pp.189-209.
Seto, K.C., Woodcock, C.E., Song, C., Huang, X., Lu, J. and Kaufmann,
R.K. (2002). Monitoring land-use change in the Pearl River Delta
using Landsat TM. Intll. Jour. Rem. Sens., v. 23(10), pp. 1985-2004.
Telesca, L. and Lasaponara, R. (2006). Quantifying intra-annual
persistent behaviour in SPOT-VEGETATION NDVI data for
Mediterranean ecosystems of southern Italy. Rem. Sens. Environ.,
v.101, pp. 95-103.
Xiao, J., Shen, Y., Ge, J., Tateishi, R., Tang, C. and Liang, Y. (2006).
Evaluating urban expansion and land use change in Shijiazhuang,
China, by using GIS and remote sensing. Landsc. Urban Plann.,
v.75 (1-2), pp. 69-80.
Yuan, F., Sawaya, K.E., Loeffelholz, B.C. and Bauer, M.E. (2005).
Landcover classification and change analysis of the Twin Cities
(Minnesota) metropolitan area by multitemporal Landsat remote
sensing. Rem. Sens. Environ., v. 98 (2-3), pp. 317-328.
Zhang, T. (2001). Community features and urban sprawl: the case
of the Chicago metropolitan region. Land Use Pol.., v.18 (3),
pp.221-232.