Acta geographica Slovenica, 58-1, 2018, 109–123
An example of illegal logging from the Municipality of Kuršumlija
in southern Serbia in 2013.
M
IT
A
R
 P
E
R
IĆ
THE USE OF NDVI AND CORINE LAND
COVER DATABESES FOR FOREST
MANAGEMENT IN SERBIA
Miomir M. Jovanović, Miško M. Milanović, Matija Zorn
58-1-Special issue_09p_818-Miomir M Jovanovic_acta49-1.qxd  12.9.2017  7:59  Page 109
Miomir M. Jovanović, Miško M. Milanović, Matija Zorn, The use of NDVI and CORINE Land Cover databases for forest …
DOI: https://doi.org/10.3986/AGS.818
UDC: 91:630*6(497.11)
630*6:528.8(497.11)
COBISS: 1.01
The use of NDVI and CORINE Land Cover databases for forest management in Serbia
ABSTRACT: This article evaluates the possible use of normalized difference vegetation index (NDVI) and
CORINE Land Cover (CLC) databases for better forest management in the municipalities of Kuršumlija
and Topola in Serbia. The forest areas obtained using CLC were up to 11.5% larger than the official forest
area estimates, whereas NDVI gave more precise results. Hence, NDVI can efficiently provide local forest
managers with essential annual information about the forest inventory.This is of a crucial importance for
preventing illegal logging, which is very prevalent in southern Serbian municipalities, which have sub-
stantial forested territory. NDVI thus very promising for Serbia and also for countries that rarely carry
out national forest inventories. This method can also easily be applied to other Balkan countries with a sim-
ilar situation regarding local forest management.
KEY WORDS: NDVI, CORINE Land Cover, forest management, illegal logging, Serbia
Raba podat kov nih zbirk NDVI in CORINE pri gos po dar je nju z goz do vi v Sr bi ji
POVZETEK: V član ku avtor ji preu ču je jo mož nost rabe podat kov nih zbirk NDVI in CORINE za bolj še gos -
po dar je nje z goz do vi v srb skih obči nah Kuršumlija in Topo la. Povr ši na goz da, ugo tov lje na z upo ra bo CLC,
je bila do 11,5 % več ja od urad no oce nje ne, med tem ko so bili rezulta ti NDVI toč nej ši. NDVI lahko lokalnim
upravljavcem gozdov zagotavlja pomembne informacije o gozdu na letni ravni. To je izjem no pomemb -
no za pre pre če va nje neza ko ni te seč nje, zna čil ne za obči ne v juž ni Srbi ji, ki so boga te z goz dom. Upo ra ba
NDVI zato obe tav na za Srbi jo in tudi dru ge drža ve, ki red ko izva jajo nacio nal ne popi se goz dov. Meto da
je pri mer na tudi za dru ge bal kan ske drža ve s po dob ni mi raz me ra mi na področ ju lokal ne ga gos po dar jenja
z goz do vi.
KLJUČNE BESEDE: NDVI, CORINE, gos po dar je nje z goz do vi, neza ko ni ta seč nja, Srbi ja
Miomir M. Jovanović, Miško M. Milanović
University of Belgrade, Faculty of Geography
miomir.m.jovanovic@gmail.com, milanovic.misko@gmail.com
Matija Zorn
Anton Melik Geographical Institute, Research Centre of the Slovenian Academy of Sciences and Arts
matija.zorn@zrc-sazu.si
The paper was submitted for publication on February 1st, 2014.
Ured niš tvo je pris pe vek pre je lo 1. fe bruar ja 2014.
110
58-1-Special issue_09p_818-Miomir M Jovanovic_acta49-1.qxd  12.9.2017  7:59  Page 110
Acta geographica Slovenica, 58-1, 2018
111
1 Introduction
Permanent clearing of forest cover was typical in the industrialized world until a few decades ago. Vast areas
of Europe and North America were cleared for industrial expansion and development of infrastructure. Today
deforestation is largely occurring in tropical countries in Africa, Asia, and Latin America (Steiningeretal. 2001;
Chowdhury 2006), as well as in taiga regions, especially in Russia (Tracy 1994; Deforestation … 2014).
Most reasons for deforestation are due to market imperfections (Jovanović 2012). Market imperfections
arise when property cannot be clearly defined, when property cannot be freely transferred, when the use
of goods cannot exclude others from such use, and when private rights cannot be protected (McKean 2000;
Tietenberg and Lewis 2012). Evidence convincingly shows that illegal and corrupt activities are a major
underlying cause of forest decline (Contreras-Hermosilla 2002; Brack 2003). The main reason for this is
that governments and private landowners cannot control these illegal operations. In addition, this lack of
control may be deliberate, is often corrupt, or may be determined by the limitations of administrative capac-
ity. One way or another, illegal use of forests is rampant (Contreras-Hermosilla 2002; Amacher et al. 2009).
Remote sensing is the detection, recognition, or evaluation of objects by means of distant sensing or
recording devices (Oštir 2006). Historically, digital remote sensing developed rapidly from aerial photography
and photo interpretation. Information extracted visually from remote sensing is widely used in forestry
(Franklin 2001; Hočevar and Kobler 2001; Hočevar and Hladnik 2006; Kobler 2012).
Given the importance and complexity of forest preservation and sustainable forest management
(Pagiola et al. 2002; Lee 2008; Ojea et al. 2012), an attempt was made to evaluate the possible use of a nor-
malized difference vegetation index (NDVI; Weier and Herring 2000) and Coordination of Information
on the Environment (CORINE) Land Cover (CORINE … 1994) in local forest management. NDVI is one
of the most widely used vegetation indices (VIs) and CORINE Land Cover (CLC) is in official use in the EU.
One of the main differences between NDVI and CLC is that, whereas NDVI focuses on the vegeta-
tion cover and its status, CLC has a much broader scope and distinguishes agricultural areas, forests and
semi-natural areas, artificial surfaces, urban fabric, industrial, commercial, and transport units, bodies of
water, wetlands, glaciers and perpetual snow, and other features (Jensen 2007).
NDVI is actually a simple graphic indicator that can be used to analyze remote sensing measurements,
whether the target observed contains live green vegetation or not (Chen 2008). NDVI was one of the most
successful of many attempts to simply and quickly identify vegetated areas and their »condition,« and it
remains the best-known and most-used index for detecting live green plant canopies in multispectral remote
sensing data (Fuller 2006; Milanović et al. 2008; Campbell and Wynne 2011; Ne Win et al. 2012). NDVI also
has the advantage of allowing comparisons between images acquired at different times (Lillesand et al. 2004).
It belongs to the VIs related to vegetation cover and its status, and it provides useful information on bio-
mass productivity and health. VIs have a direct correlation with leaf chlorophyll content and leaf area index
(LAI) and vary in relation to vegetation cycle and phenology (Vohlandetal. 2007; Montandon and Small 2008).
They are also sensitive to other external factors, such as the contribution of the soil and background opti-
cal behavior where the vegetation does not completely cover the ground, the geometry of view due to sensor
angle of acquisition and to Sun position, atmospheric effects, and other factors (Franklin 2001; De Jong
and Van der Meer 2005; Jensen 2007; Campbell and Wynne 2011).
NDVI, like all VIs, relates the spectral absorption of chlorophyll in the red with a reflection phenomenon
in the near infrared, influenced by the leaf structure type (Wang and Tenhunen 2004).
In contrast, CLC is a European program launched in 1985 by the European Commission, aimed at obtain-
ing a unique and comparable dataset of land cover for Europe. The aim of CLC is to gather information
related to the environment on certain priority topics for the European Union: air, water, soil, land cover,
coastal erosion, biotopes, and so on. The main goals of the CLC program are to acquire information about
the environment to address the European Community policy, to assess the effectiveness of legislation, to
integrate environmental and political aspects, to unify heterogeneous thematic cartographies of Europe
at various levels (international, national, regional, local), and to update data at regular intervals, every five
to ten years (Bossard et al. 2000; Neumann et al. 2007). CLC is a map of the European environmental land-
scape based on interpretation of satellite images.
The data have been validated using local cartography and ground surveys (Heymannetal. 1994; Perdigão
and Annoni 1997; Genovese et al. 2001). CLC also has an NDVI module for creating vegetation maps, but
the deviations in its final results are substantial due to the highly inappropriate scale of Serbian data.
58-1-Special issue_09p_818-Miomir M Jovanovic_acta49-1.qxd  12.9.2017  7:59  Page 111
Miomir M. Jovanović, Miško M. Milanović, Matija Zorn, The use of NDVI and CORINE Land Cover databases for forest …
For creating CLC maps for the municipalities of Kuršumlija and Topola, image processing was car-
ried out and a digital elevation model was made based on the municipalities' boundaries and Landsat satellite
color composites, and a pseudo-color composite with bands 4, 5, 1 and adequate contrast was applied. Datasets
and maps for Serbia, mainly IMAGE2000 and CLC2000  class, were extracted from the European
Environmental Agency (EEA) website, with a transfer data scale of 1 : 100,000, which resulted in a very
high level of imprecision (Büttner and Kleeschulte 2005).
Of particular interest to this study is that the smallest unit is 25 ha in the original CLC project,
although a recent approach yields more precise results because changes < 25 ha and > 5 ha are mapped
(CLC2006 … 2007). Nevertheless, even the smallest 5 ha areas, which are highly appropriate at the EU scale,
do not properly reflect the land-use situation at the local scale in a country where landscapes and land-use
change across very short distances (Hočevar and Kobler 2001; Gabrovec and Petek 2004).
This article shows that remote sensing data collection and analysis methods have great importance
for local forest management in Serbia. In Serbia around 30% of land is forested (of which 50% is state-owned
forests and 50% privately owned). Forest management (of both privately owned and state-owned forests)
is also very poor (Forestry … 2006).
The purpose of this article is to improve the local forest management system in Serbia through more
precise methods for assessing land-use changes in forest areas. The study evaluated NDVI and CLC, which
are viewed as very efficient tools for classifying and estimating different land cover types of large and remote
areas (Meng et al. 2009). Although they both proved to be very effective in the EU, CLC is mostly used as
a regional database. Nonetheless, in Serbia they both (recently) became very popular tools for studies at
the local level (Report … 2009). This article shows that they are not equally effective at the local level in
the Serbian context.
112
LEGEND
Municipality of Topola2
1 Municipality of Kuršumlija
Content by: Miško Milanović
Map by: Miško Milanović
© 2014, University of Belgrade, Faculty of Geography
EUROPE
SERBIA
1
2
Serbia
Kosovo
Figure 1: Location of the municipalities of Kuršumlija and Topola (Serbia).
58-1-Special issue_09p_818-Miomir M Jovanovic_acta49-1.qxd  12.9.2017  7:59  Page 112
2 Materials and methods
In the study it was not possible to make a reliable long-term comparative analysis between NDVI and CLC
data and official forest inventories because national forest inventories have very rarely been carried out
in Serbia. Such inventories were carried out at roughly twenty-year intervals: in 1961, 1979, and 2003–2006.
Since 2007, official estimates of forest areas have been made annually.
The study was carried out for the municipalities of Topola and Kuršumlija (Figure 1). Data obtained
using NDVI and CLC for spring/summer 2006 were analyzed and compared to official forest area esti-
mates for 2006 created at the end of the same year. The Municipality of Topola is located in central Serbia,
and the Municipality of Kuršumlija lies in southern Serbia.
NDVI and CLC data for both municipalities are based on Landsat 5 Thematic Mapper (TM) satellite
images (Figures 2 and 3) for 2006, which were created during spring/summer (August), with minimum
clouds (10 to 20%; Chavez 1996). In order to remove atmospheric effects from the NDVI final results, Idrisi
software was used for data preprocessing.
For calculating NDVI, satellite (Landsat) imagery (which has a resolution of approximately 30 m) and
pan-sharpening images (with 15 m resolution) were used to obtain more precise results.
Acta geographica Slovenica, 58-1, 2018
113
0 5 10 km2.5
Content by: Miško Milanović
Map by: Miško Milanović
Source: U.S. Geological Survey, 2014
© 2014, University of Belgrade, Faculty of Geography
Figure 2: The Municipality of Kuršumlija.
L
A
N
D
S
A
T
 I
M
A
G
E
58-1-Special issue_09p_818-Miomir M Jovanovic_acta49-1.qxd  12.9.2017  7:59  Page 113
Miomir M. Jovanović, Miško M. Milanović, Matija Zorn, The use of NDVI and CORINE Land Cover databases for forest …
NDVI was used and necessary corrections/transformations were applied for visible red in constellation
with the infrared spectrum of satellite images using the following procedure: GIS Analysis / Mathematical
Operation / Image Calculator, and then the equation NDVI = (NIR – RED) / (NIR + RED), in which NIR
is the near-infrared channel and RED is the red channel from the visible part of the spectrum (Hájek 2008;
Johnson and Trout 2012).
Basic tasks included analysis and photo interpretation of elements, occurrences, and processes detect-
ed on images using specialized GIS software (Idrisi 15-Andes) for processing remotely sensed images through
application of NDVI.
Shadows can cause NDVI values to be lower than their actual values. In this sense, »empirical topo-
graphic corrections have proven only marginally successful« (Franklin 2001). Because shadow areas were
less than 5% in the Municipality of Kuršumlija and less than 3% in the Municipality of Topola, no topo-
graphic corrections were made.
Characteristic NDVI signatures are as follows: NDVI of dense vegetation canopy tends to have positive
values (0.3 to 0.8); clouds and snowfields are characterized by negative values of this index; bodies of water
(e.g., oceans, seas, lakes, and rivers) has rather low reflectance in both spectral bands (at least away from shores),
thus resulting in very low positive or even slightly negative NDVI values; soils generally exhibit a near-infrared
spectral reflectance somewhat larger than the red, and thus also tend to generate rather small positive NDVI
values (0.1 to 0.2); very low values of NDVI (0.1 and below) correspond to barren areas of rock, sand, or
snow; moderate values represent shrub and grassland (0.2 to 0.3); and high values (0.6 to 0.8) indicate tem-
114
L
A
N
D
S
A
T
 I
M
A
G
E
Content by: Miško Milanović
Map by: Miško Milanović
Source: U.S. Geological Survey, 2014
© 2014, University of Belgrade, Faculty of Geography
0 5 10 km2.5
Figure 3: The Municipality of Topola.
58-1-Special issue_09p_818-Miomir M Jovanovic_acta49-1.qxd  12.9.2017  7:59  Page 114
Acta geographica Slovenica, 58-1, 2018
115
0.3
0.4
0.5
0.6
0.7
NDVI Index
0 2 4 6 8 10 km
Content by: Miško Milanović
Map by: Miško Milanović
© 2014, University of Belgrade, Faculty of Geography
Figure 4: Vegetation cover in the Municipality of Kuršumlija for 2006 obtained from NDVI.
perate and tropical rainforests (Finelli et al. 1996; Schmitt and Ruppert 1996). Negative values of NDVI rang-
ing from 0 to –0.3 are displayed in shades from light green to dark purple. These low negative values are
detected in arable agricultural land (without vegetation) and are shown in shades of light green. On the other
hand, vegetation areas are presented with values between 0 and 1. Grassy areas, meadows, and pastures have
values that range from zero (in yellow, due to more intense reflectance of infrared radiation) up to 0.13 (light
orange tones). Shrub vegetation has an NDVI value of 0.25 because reflectance of infrared rays decreases
(darker red tones). Forest vegetation, with maximal positive NDVI values of 0.85 (due to minimal reflectance
of infrared rays), is easily observed. Coniferous forest has an NDVI value above 0.5, mixed forest between
0.35 and 0.5, and broad-leaved forest between 0.3 and 0.4 (Bakx 1995; De Jong and Van der Meer 2005).
3 Results
After image processing it was determined (Table 1) that forest areas encompass 529.83 km2 or 55.7% of
the total area of the Municipality of Kuršumlija, much higher than the average 30% for Serbia; and 50.73 km2
or 14.2% of the total area of the Municipality of Topola, which is approximately half of the Serbian average.
58-1-Special issue_09p_818-Miomir M Jovanovic_acta49-1.qxd  12.9.2017  7:59  Page 115
Miomir M. Jovanović, Miško M. Milanović, Matija Zorn, The use of NDVI and CORINE Land Cover databases for forest …
When these NDVI results are compared with official forest area estimates for the same year (2006)
(Municipalities … 2007), there is only +0.12 km2 of difference for Topola's forest area, and –27 km2 differ-
ence for Kuršumlija (Table 3).
Figures 4 and 5 present a raster of NDVI (NIR band and RED band) from Landsat 5 TM (bands 4, 5, 1)
satellite images. The images were created in August 2006. Figures 6 and 7 present vegetation cover obtained
from CLC.
When the (latest available) CLC results for 2006 were compared with official forest area estimates for
the same year (Tables 1–3), some inconsistencies became apparent:
• The total areas for the municipalities of Kuršumlija and Topola obtained from CLC were smaller than
the official forest statistics: instead of 952 km2 only 942.9 km2 for Kuršumlija, and instead of 356 km2
only 348.9 km2 for Topola;
• Forest areas obtained from CLC were up to  11.5% larger than the official forest area estimates.
Kuršumlija's forest area obtained from CLC (630.45 km2) is 26 km2 larger than the official forest area
estimates (604.41km2) for this municipality (+4.3%), and Topola's forest area obtained from CLC (58km2)
is 6 km2 larger than the official forest area estimates (52 km2, + 11.5%).
116
Content by: Miško Milanović
Map by: Miško Milanović
© 2014, University of Belgrade, Faculty of Geography
0 2 4 6 8 10 km
NDVI Index
1
–1
Figure 5: Vegetation cover in the Municipality of Topola for 2006 obtained from NDVI.
58-1-Special issue_09p_818-Miomir M Jovanovic_acta49-1.qxd  12.9.2017  7:59  Page 116
Acta geographica Slovenica, 58-1, 2018
117
0 2 4 6 8 10 km
Content by: Miško Milanović
Map by: Miško Milanović
© 2014, University of Belgrade, Faculty of Geography
Settlements
Non-irrigated arable land
Complex cultivation patterns
Land principally occupied
by agriculture, with signi$cant areas
of natural vegetation
Natural grasslands
Pastures
Broad/leaved forest
Coniferous forest
Mixed forest
Transitional woodland-shrub
Sparsely vegetated areas
LEGEND
Figure 6: Vegetation cover in the Municipality of Kuršumlija for 2006 obtained from CLC.
Table 1: Vegetation cover in the municipalities of Kuršumlija and Topola for 2006 obtained from NDVI.
Land cover Kuršumlija Topola
(km2) (%) (km2) (%)
Broad-leaved forest 562.71 59.10 49.40 13.84
Coniferous forest 6.46 0.68 0.62 0.17
Mixed forest 8.23 0.86 2.10 0.59
Pastures 32.20 3.40 – –
Transitional woodland-shrub 51.55 5.41 – –
Sparsely vegetated areas 9.48 0.99 – –
Land principally occupied by agriculture,
with significant areas of natural vegetation – – 63.25 17.72
Other 281.37 29.56 241.63 67.68
Total 952.00 100.00 357.00 100.00
58-1-Special issue_09p_818-Miomir M Jovanovic_acta49-1.qxd  12.9.2017  7:59  Page 117
Miomir M. Jovanović, Miško M. Milanović, Matija Zorn, The use of NDVI and CORINE Land Cover databases for forest …
118
0 2 4 6 8 10 km
Content by: Miško Milanović
Map by: Miško Milanović
© 2014, University of Belgrade, Faculty of Geography
Boundary
Settlements
Green urban areas
Non–irrigated arable land
Natural grasslands
Complex cultivation patterns
Land principally occupied by agriculture,
with signi$cant areas of natural vegetation
Vineyards
Broad-leaved forest
Coniferous forest
Mixed forest
LEGEND
Transitional woodland-shrub
Pastures
Figure 7: Vegetation cover in the Municipality of Topola for 2006 obtained from CLC.
Table 2: Land cover in the municipalities of Kuršumlija and Topola for 2006 obtained from CLC.
Land cover Kuršumlija Topola
(km2) (%) (km2) (%)
Settlements 4.60 0.49 9.11 2.61
Green urban areas – – 0.82 0.23
Non-irrigated arable land 0.42 0.04 36.51 10.46
Natural grasslands 25.74 2.73 14.44 4.14
Complex cultivation patterns 78.73 8.35 154.94 44.41
Land principally occupied by agriculture,
with significant areas of natural vegetation 102.25 10.84 73.17 20.97
Broad-leaved forest 620.68 65.82 55.68 15.96
Coniferous forest 3.63 0.38 0.19 0.05
Mixed forest 6.14 0.65 2.12 0.61
Pastures 24.18 2.56 1.11 0.32
Transitional woodland-shrub 75.78 8.04 0.81 0.23
Sparsely vegetated areas 0.78 0.08 – –
Total 942.93 100.00 348.90 100.00
58-1-Special issue_09p_818-Miomir M Jovanovic_acta49-1.qxd  12.9.2017  7:59  Page 118
4 Discussion
Although both CLC and NDVI have recently been used in Serbia for studies at the local level, the main
problem with CLC data is that: a) although CLC data are produced at various levels (international, nation-
al, regional, and local; Bossard et al. 2000; Neumann et al. 2007), CLC is actually a predominantly regional
database, updated rarely (every five to ten years), whereas NDVI is available for every year; and b) NDVI
is much more precise than CLC.
When official statistics and NDVI and CLC forest areas were compared for the same year (2006), NDVI
was more precise than CLC. Because both NDVI and CLC used the same Landsat satellite images and the
same (NDVI) methodology, these major differences in the data obtained were due to the different spatial
resolution of NDVI and CLC.
Whereas CLC does not go below the range of 4 to 5 ha, NDVI easily deals with minimum space units
of 25 m2. This proved to be decisive for Serbia, where privately owned forest parcels, which account for
half of the total forest area of the country, usually cover much smaller areas (the average private holding
is 0.5 ha; Glavonjić et al. 2005). In short, CLC proved not to be very suitable for local forest management
in Serbia (questionable results regarding forests were also determined in Slovenia; e.g., Gabrovec and
Petek 2004). In addition, apart from the obvious CLC imprecision for studies at the local level, CLC data
are not available for every year.
When compared with official forest area estimates, the NDVI results show a mere +0.12 km2 (+0.2%)
difference for the Municipality of Topola's forest area, and a –27.01km2 (–4.7%) difference for the Municipality
of Kuršumlija (Table 3). Not only do these results completely fit within the ± 5% margin of error allowed
for this method (Eastman 2001; Lunetta et al. 2007), but they also allow room for further analysis and inves-
tigation.
Because the NDVI aerial photos were taken during spring/summer, whereas official forest area esti-
mates are made at the end of the year, NDVI values would be expected to be higher, not lower–at least for
the Municipality of Kuršumlija (known for its illegal logging). Moreover, because additional NDVI for-
est area estimates were made for 2011 (Table 4), it seems that even for 2006 this study's NDVI results better
fit the forest area trajectory of Kuršumlija for the 2006–2011 period than do the official statistics (the offi-
cial forest inventory for 2006 is 604.41 km2 and NDVI results 577.4 km2; and the official forest inventory
for 2011 is 544.3 km2 and NDVI results 529.8 km2).
Acta geographica Slovenica, 58-1, 2018
119
Table 3: Forest areas according to official statistics and calculated on the basis of NDVI and CLC for 2006.
Municipality Municipality: Forest area NDVI – official CLC – official
total area Official statistics Calculated on the Calculated on the statistics difference statistics difference
(km2) (km2)* basis of NDVI basis of CLC (km2) (km2)
(km2) (km2)
Topola 356 52.00 52.12 57.99 +0.12 +5.99
Kuršumlija 952 604.41 577.40 630.45 –27.01 +26.04
*Source: Municipalities…2008.
Table 4: Forest areas according to official statistics and calculated on the basis of NDVI for 2011 and CLC for 2006.
Municipality Municipality: Forest area NDVI – official CLC – official
total area Official statistics Calculated on the Calculated on the statistics difference statistics difference
(km2) (km2)* basis of NDVI basis of CLC (km2) (km2)
(km2) (km2)
Topola 357 52.0494 50.73 57.99 –1.3194 +5.9407
Kuršumlija 952 544.2856 529.83 630.45 –14.4556 +86.1644
*Source: Municipalities…2013.
58-1-Special issue_09p_818-Miomir M Jovanovic_acta49-1.qxd  12.9.2017  7:59  Page 119
Miomir M. Jovanović, Miško M. Milanović, Matija Zorn, The use of NDVI and CORINE Land Cover databases for forest …
The main reason that the (slightly smaller) NDVI results possibly better fit the forest area trajectory of
Kuršumlija than the official inventory is that this municipality is known for illegal logging. According to the
state-owned forest-management company Srbijašume, in Kuršumlija more than 40,000 m3 of timber was
illegally cut during the last thirteen years alone, and that municipality also experienced a 10% loss in for-
est area in the last few years alone (Forestry…2006; Anfodilloetal. 2008; Illegal…2009, Municipalities…2013).
Obviously, governments often cannot efficiently control these illegal operations. As Contreras-Hermosilla
(2000) points out: »This lack of control can be either deliberate, often corrupt, or determined by the limita-
tions of administrative capacity. One way or the other, illegal use of forests is rampant in most forested countries.
By their very nature, the true extent of illegal operations in the forestry sector cannot be known with preci-
sion, but evidence suggests that such activities are important and that they constitute an important underlying
cause of forest decline.«
Because this research strongly implies that illegal logging in Kuršumlija is not properly covered by cur-
rent official forest area estimates, further NDVI research on the extent of illegal logging in southern Serbian
municipalities is of the utmost importance.
In short, because the Municipality of Kuršumlija has a large territory (952 km2), with more than 544 km2
(or 55.7%) of its total area covered by forests, and because NDVI can be performed very quickly, it is obvi-
ous that NDVI can provide local forest managers in Kuršumlija with much essential annual information
about the forest inventory (Chen et al. 2004; Bellone 2009; Fensholt et al. 2009; Martinez and Gilabert 2009;
Alessandrini et al. 2010; Corral-Rivas 2010). This is of crucial importance for preventing illegal logging,
which is very prevalent in this southern Serbian municipality (Forestry … 2006; Anfodillo et al.  2008;
Illegal … 2009).
5 Conclusion
Despite certain shortcomings (Franklin 2001; Campbell and Wynne 2011), classification and area estimation
of various land-cover types based on remote sensing has obviously advanced to a point where it surpasses
old wood inventory techniques, especially in the case of Serbia. Specifically:
• It is relatively cheap and quick, and it can provide forest managers with essential information;
• It is easy to implement, which is of crucial importance for Serbia, where national forest inventories have
been carried out very rarely. The last three national forest inventories were carried out at roughly twen-
ty-year intervals; however, since the last national forest inventory (2003–2006), necessary updates have
been made every year, but only at the municipality level;
• The objectivity of these methods can significantly help in avoiding corruption in forest management
because corruption is one of the main weaknesses of Serbia's economy.
Through this analysis of NDVI and CLC results for the municipalities of Kuršumlija and Topola, CLC
was shown not to be a very suitable tool for local forest management in Serbia. On the other hand, it is
evident that NDVI, especially in southern Serbian municipalities with prevalent illegal logging (like
Kuršumlija), can provide local forest managers with much annual information about forest areas. This is
of crucial importance for monitoring (and consequently preventing) illegal logging.
NDVI is also very promising for countries like Serbia, which very rarely carry out national forest inven-
tories. It is easy implemented and it has objectivity that can greatly help avoid corruption and illegal logging
in forest management.
ACKNOWLEDGEMENT: This work was supported by the Ministry of Science and Technological
Development of the Republic of Serbia under grant no. 37010.
6 References
Alessandrini, A., Vessella, F., Di Filippo, A., Salis, A., Santi, L., Schirone, B., Piovesan, G. 2010: Combined
dendroecological and normalized difference vegetation index analysis to detect regions of provenance
in forest species. Scandinavian Journal of Forest Research  25-8. DOI: http://dx.doi.org/10.1080/
02827581.2010.485776
120
58-1-Special issue_09p_818-Miomir M Jovanovic_acta49-1.qxd  12.9.2017  7:59  Page 120
Amacher, G., Ollikainen, M., Koskela, E. 2009: Economics of forest resources. Cambridge.
Anfodillo, T., Carrer, M., Giacoma, E., Lamedica, S., Pettenella, D. 2008: Challenges and priority for adapt-
ing the management of Carpathians forests to new environmental and socio-economic conditions. Internet:
http://www.carpathianconvention.org/tl_files/carpathiancon/Downloads/03%20Meetings%20and%
20Events/COP/2008_COP2_Bucharest/COP2_Information%20documents/16Report2_ForestPolicy.pdf
(17. 4. 2014).
Bakx, P. 1995: Digital image processing of remotely sensed data. Enschede.
Bellone, T., Boccardo, P., Perez, F. 2009: Investigation of vegetation dynamics using long-term normalized
difference vegetation index time-series. American Journal of Environmental Sciences 5-4. New York.
DOI: http://dx.doi.org/10.3844/ajessp.2009.461.467
Bossard, M., Feranec, J., Otahel, J. 2000: CORINE Land Cover technical guide – addendum 2000. Technical
report 40. Copenhagen. Internet: http://www.eea.europa.eu/publications/tech40add (17. 4. 2014).
Brack, D. 2003: Illegal logging and the illegal trade in forest and timber products. International Forestry
Review 5-3. Oxford. DOI: http://dx.doi.org/10.1505/IFOR.5.3.195.19148
Büttner, G., Kleeschulte, S. 2005: Implementation of CLC2000 in Serbia-Montenegro – technical annex.
Campbell, J., Wynne, R. 2011: Introduction to remote sensing. New York.
Chakravarty, S., Ghosh, S. K., Suresh, C. P., Dey, A. N., Shukla, G. 2012: Deforestation: causes, effects and
control strategies. Global Perspectives on Sustainable Forest Management. Rijeka.
Chavez, S. jr. 1996: Image-based atmospheric corrections – revisited and improved. Photogrammetric
Engineering and Remote Sensing 62.
Chen, C. H. (ed.) 2008: Image processing for remote sensing. New York.
Chen, J., Jonsson, P., Tamura, M., Gu, Z., Matsushita, B., Eklundh, L. 2004: A simple method for recon-
structing a high-quality NDVI time-series data set based on the Savitzky–Golay filter. Remote Sensing
of Environment 91, 1–2. DOI: http://dx.doi.org/10.1016/j.rse.2004.03.014
Chowdhury, R. R. 2006: Driving forces of tropical deforestation: The role of remote sensing and spatial
models. Singapore Journal of Tropical Geography 27-1. DOI: http://dx.doi.org/10.1111/j.1467-9493.2006.00241.x
CLC2006 Technical Guidelines. EEA Technical report No.17. Copenhagen, 2007. Internet: http://www.eea.europa.eu/
publications/technical_report_2007_17 (17. 4. 2014).
Contreras-Hermosilla, A. 2000: The underlying causes of forest decline. Center for International Forestry
Research occasional paper 30. Bogur. Internet: http://www.cifor.org/publications/pdf_files/OccPapers/
OP-030.pdf (17. 4. 2014).
CORINE Land cover  – Part  1: Methodology. European Environment Agency,  1994. Internet:
http://www.eea.europa.eu/publications/COR0-part1 (17. 4. 2014).
Corral-Rivas, J. J., Wehenkel, C., Castellanos-Bocaz, H. A., Vargas-Larreta, B., Dieguez-Aranda, U. 2010:
A permutation test of spatial randomness: Application to nearest neighbour indices in forest stands.
Journal of Forest Research 15-4. DOI: http://dx.doi.org/10.1007/s10310-010-0181-1
Dale, V. H. 2008: Management of forests as ecosystems: A success story or a challenge ahead? Success, lim-
itations, and frontiers in ecosystem Science. New York.
Deforestation in the Russian Taiga. Internet: http://www-pub.naz.edu:9000/~jwitten2/about2.htm (17.4.2014).
De Jong, S. M., Van der Meer, F. D. 2005: Remote sensing image analysis: including the spatial domain.
Dordrecht.
Eastman R. J. 2001: IDRISI32 Release 2 – Tutorial. Worcester.
Fensholt, R., Rasmussen, K., Nielsen, T. T., Mbow, C. 2009: Evaluation of earth observation based long term
vegetation trends – Intercomparing NDVI time series trend analysis consistency of Sahel from
AVHRR GIMMS, Terra MODIS and SPOT VGT data. Remote Sensing of Environment 113-9.
DOI: http://dx.doi.org/10.1016/j.rse.2009.04.004
Finelli, M., Gelli, G., Poggi, G. 1996: Multispectral-image coding by spectral classification. Proceedings:
International conference on image processing 2. Lausanne.
Forestry development strategy for Serbia. Ministry of Agriculture, Forestry and Water Management,
Directorate of Forests. Belgrade, 2006.
Franklin, S. 2001: Remote sensing for sustainable forest management. New York.
Gabrovec, M., Petek, F. 2004: V Sloveniji vedno manj gozda? (Ne)uporabnost metodologije CORINE Land
Cover. Delo 46-303 (30. 12. 2004). Ljubljana.
Acta geographica Slovenica, 58-1, 2018
121
58-1-Special issue_09p_818-Miomir M Jovanovic_acta49-1.qxd  12.9.2017  7:59  Page 121
Miomir M. Jovanović, Miško M. Milanović, Matija Zorn, The use of NDVI and CORINE Land Cover databases for forest …
Genovese, G., Vignolles, C., Negre, T., Passera, G. 2001: A methodology for a combined use of normalised
difference vegetation index and CORINE land cover data for crop yield monitoring and forecasting:
A case study on Spain. Agronomie 21-1. DOI: http://dx.doi.org/10.1051/agro:2001111
Glavonjić, B., Jović, D., Vasiljević, A., Kankaras, R. 2005: Forest and forest products country profile: Serbia
and Montenegro. Geneva Timber and Forest Discussion Paper 40. Geneva. Internet: http://www.unece.org/
fileadmin/DAM/timber/docs/dp/dp-40.pdf (17. 4. 2014).
Hájek, F. 2008: Process-based approach to automated classification of forest structures using medium for-
mat digital aerial photos and ancillary GIS information. European Journal of Forest Research 127-2.
DOI: http://dx.doi.org/10.1007/s10342-007-0188-0
Heymann, Y., Steenmans, C. H., Croissille, G., Bossard, M. 1994: CORINE land cover, technical guide.
Luxembourg.
Hočevar, M., Kobler, A. 2001: Uporaba podatkov CORINE Land Cover za izgradnjo informacijskega sis-
tema o gozdovih na državni ravni. Gozdarski vestnik 46-3.
Hočevar, M., Hladnik, D. 2006: Development of forest monitoring methods for sustainable forest man-
agement in Slovenia. Nature-based forestry in Central Europe: alternatives to industrial forestry and
strict preservation. Ljubljana.
Illegal logging activities in the Republic of Serbia – A. Fact-finding study. Regional Environmental Center.
Belgrade, 2009. Internet: http://illegallogging.rec.org/publications/SRB_IL%20Fact%20Finding_Final.pdf
(17. 4. 2014).
Jensen, J. R. 2007: Remote sensing of the environment: an earth resource perspective. Upper Saddle River.
Johnson, L. F., Trout, T. J. 2012: Satellite NDVI assisted monitoring of vegetable crop evapotranspiration
in California's San Joaquin Valley. Remote Sensing 4-2. DOI: http://dx.doi.org/10.3390/rs4020439
Jovanović, M. 2012: Kuznets curve and urban transport – The scope of I+M programs. Glasnik Srpskog
geografskog društva 92-4. DOI: http://dx.doi.org/10.2298/GSGD1204127J
Klaučo, M., Weis, K., Stankov, U., Arsenovic, D., Marković, V. 2012: Ecological significance of land-cover
based on interpretation of human-tourism impact. A case from two different protected areas (Slovakia
and Serbia). Carpathian Journal of Earth and Environmental Sciences 7-3.
Kobler, A. 2012: Daljinsko zaznavanje gozdov z lidarjem. Ljubljana.
Lee, B., Kim, S. Y., Chung, J., Park, P. S. 2008: Estimation of fire severity by use of Landsat TM images and its
relevance to vegetation and topography in the 2000 Samcheok forest fire. Journal of Forest Research 13.
DOI: http://dx.doi.org/10.1007/s10310-008-0072-x
Lillesand, T. M., Kieffer, R. W., Chipman, J. W. 2004: Remote sensing and image interpretation. New York.
Lisec, A., Pišek, J., Drobne, S. 2013: Suitability analysis of land use records of agricultural and forest land for
detecting land use change on the case of the Pomurska Statistical Region. Acta geographica Slovenica 53-1.
DOI: http://dx.doi.org/10.3986/AGS53104
Lunetta, R. S., Knight, J. F., Ediriwickrema, J., Lyon, J. G., Worthy, L. D. 2007: Land-cover change detection
using multi-temporal MODIS NDVI data. Remote Sensing of Environment 105-2. DOI: http://dx.doi.org/
10.1016/j.rse.2006.06.018
Martínez, B., Gilabert, M. A. 2009: Vegetation dynamics from NDVI time series analysis using the wavelet
transform. Remote Sensing of Environment 113-9. DOI: http://dx.doi.org/10.1016/j.rse.2009.04.016
McKean, M. 2000: Common Property: What is it, what is it good for, and what makes it work? People and
forests: communities, institutions, and governance. Cambridge.
Meng, Q., Cieszewski, C., Madden, M. 2009: Large area forest inventory using Landsat ETM+: A geosta-
tistical approach. ISPRS Journal of Photogrammetry and Remote Sensing 64-1. DOI: http://dx.doi.org/
10.1016/j.isprsjprs.2008.06.006
Milanović, M., Lješević, M., Bakrač, S., Novković, I. 2008: Possibility analysis of vegetation processing satel-
lite images of infrared spectrum. Proceedings of the 9th International Conference on Fundamental and
Applied Aspects of Physical Chemistry 2. Belgrade.
Montandon, L. M., Small, E. E. 2008: The impact of soil reflectance on the quantification of the green veg-
etation fraction from NDVI. Remote Sensing of Environment 112-4. DOI: http://dx.doi.org/10.1016/
j.rse.2007.09.007
Municipalities and regions of the Republic of Serbia. Statistical Office of the Republic of Serbia. Belgrade, 2008.
Municipalities and regions of the Republic of Serbia. Statistical Office of the Republic of Serbia. Belgrade, 2013.
122
58-1-Special issue_09p_818-Miomir M Jovanovic_acta49-1.qxd  12.9.2017  7:59  Page 122
Ne Win, R., Suzuki, R., Takeda, S. 2012: Remote sensing analysis of forest damage by selection logging in
the Kabaung Reserved Forest, Bago Mountains, Myanmar. Journal of Forest Research 17-2. DOI:
http://dx.doi.org/10.1007/s10310-011-0276-3
Neumann, K., Herold, M., Hartley, A., Schmullius, C. 2007: Comparative assessment of CORINE2000 and
GLC2000: Spatial analysis of land cover data for Europe. International Journal of Applied Earth
Observation and Geoinformation 9-4. DOI: http://dx.doi.org/10.1016/j.jag.2007.02.004
Ojea, E., Ruiz-Benito, P., Markandya, A., Zavala, M. A. 2012: Wood provisioning in Mediterranean forests:
A bottom-up spatial valuation approach. Forest Policy and Economics 20. DOI: http://dx.doi.org/10.1016/
j.forpol.2012.03.003
Oštir, K. 2006: Daljinsko zaznavanje. Ljubjana.
Pagiola, S., Landell-Mills, N., Bishop, J. 2002: Selling forest environmental services. London.
Perdigão, V., Annoni, A. 1997: Technical and methodological guide for updating CORINE land cover data
base. Internet: http://www.ec-gis.org/docs/F27057/CORINE.PDF (17. 4. 2014).
Report on the status of land in the Republic of Serbia. Agency for environmental protection, Ministry of
Agriculture and Environmental Protection, Belgrade, 2009.
Schmitt, U., Ruppert, G. 1996: Forest classification of multispectral mosaicking satellite images. Archives
of Photogrammetry and Remote Sensing 31.
Skudnik, M., Kušar, G. 2013: Use of 3d process-based model to determine forests protecting against rock-
fall – case study Kamniška Bistrica. Acta geographica Slovenica 51-2. DOI: http://dx.doi.org/10.3986/
AGS51202
Steininger, M. K., Tucker, C. J., Townshend, J. R. G., Killeen, T. J., Desch, A., Bell, V., Ersts, P. 2001: Tropical
deforestation in the Bolivian Amazon. Environment Conservation Journal 28-2. DOI: http://dx.doi.org/
10.1017/S0376892901000133
Tietenberg, T., Lewis, L. 2012: Environmental and natural resource economics. Boston.
Tracy, L. A. 1994: The deforestation of Siberia: economic and environmental problems in Russian Forest
Management. Forest Industry Lecture 32.
Vohland, M., Stoffels, J., Hau, C., Schüler, G. 2007: Remote sensing techniques for forest parameter assess-
ment: multispectral classification and linear spectral mixture analysis. Silva Fennica 41-3.
Wang, Q., Tenhunen, J. 2004: Vegetation mapping with multitemporal NDVI in North Eastern China Transect
(NECT). International Journal of Applied Earth Observation and Geoinformation 6. DOI: http://dx.doi.org/
10.1016/j.jag.2004.07.002
Weier, J., Herring, D. 2000: Measuring vegetation (NDVI & EVI). Earth Observatory, 30. 8. 2000. Internet:
http://earthobservatory.nasa.gov/Features/MeasuringVegetation/measuring_vegetation_1.php (4.11.2012).
Acta geographica Slovenica, 58-1, 2018
123
58-1-Special issue_09p_818-Miomir M Jovanovic_acta49-1.qxd  12.9.2017  7:59  Page 123