Tomáš Václavík
Research Assistant and PhD student of Geography and Urban Regional Analysis
Center for Applied Geographic Information Science (CAGIS)
Department of Geography and Earth Sciences
University of North Carolina at Charlotte
9201 University City Boulevard, Charlotte, NC 28223-0001, USA
Email: [email protected]
Tel.: +1-347-272-2825
URL: http://webpages.uncc.edu/~tvaclavi/ or http://tova.euweb.cz/
IDENTIFYING TRENDS IN LAND USE/LAND COVER
CHANGES IN THE OLOMOUC REGION, CZECH REPUBLIC
ABSTRACT
The Olomouc region in the Czech Republic has undergone significant changes in the past several decades, such as the change in a political system of the country in 1989. Although the political and cultural transformation is generally recognized as an important driver of land use
(Ptáček 2000), there have been few studies conducted that would empirically assess and quantify land use/land cover changes in the Czech Republic, especially in the context of the post- socialistic transformation (Fanta et al. 2004; Zemek et al. 2005). In this study, I present an approach for identifying major land use/land cover changes in the Olomouc region applying
1 remote sensing techniques to compare data from multispectral satellite sensors acquired twelve years before and twelve years after the revolution in 1989. I pay closer attention to specific trends in land cover changes: changes in agricultural areas, forest areas, and residential development.
The results support initial assumptions that the land cover will reflect the changes in human perception of landscape and natural resources, such as smaller need for intensive agriculture, shift to environmental friendly management of forested areas, or increasing development and suburbanization.
1. INTRODUCTION
The Czech Republic is currently undergoing transformation from the centralized regime of communist dictatorship towards modern democratic state. The communist collectivization of land in 1950s, the reintroduction of democracy and market economy after the revolution in 1989 and the preparation of the Czech Republic for ingression to the European Union in 2004 are recognized as main drivers of land use in the last 50 years (Fanta et al. 2005). This research pays closer attention to specific trends in land use changes within the last 25 years: changes in agricultural areas, forest areas, and residential development.
The main objective of this study is to analyze relevant remote sensing data from 1976 and
2001 and identify the locations, types, and trends of the main land cover changes in the last 25 years. I assume that the land cover will reflect changes in human perception of landscape and natural resources, such as decreased need for intensive agriculture (Fanta et al. 2004), shift to environmentally friendly management of forested areas considering natural species composition and potential vegetation (Neuhäuslová 1998), or increasing development and suburbanization
(Ptáček 1998; Jackson 2002).
2 2. METHODS
2.1 Study Site
The study area chosen for this project is the Olomouc region in the eastern Moravian part of the Czech Republic (Figure 1). The study area of 5012 km2 covers the majority of the
Olomouc county administration unit. The central part is formed by a wide alluvial plane of the upper stream of the Morava River and is protected as the Litovelske Pomoravi Protected
Landscape Area. The central part is surrounded by undulated hills of the Zabrezska and
Drahanska uplands from the west, and Nizky Jesenik mountain range from the northeast. The lowland areas are highly urbanized and include substantive agricultural areas. Major forested habitats of the region are located in northeastern uplands, predominantly composed of coniferous and mixed stands, and used for timber production.
Litovelske Pomoravi Protected Landscape Area
Czech Republic
Olomouc region o
Figure 1: Study area
2.2 Data Collection
The Landsat Multispectral Scanner and Enhanced Thematic Mapper scenes were acquired
3 for change detection analysis (Table 1). The MSS data included one scene (path 204, row 25) from 8 May 1976; the ETM+ data included two scenes (path 190, row 25 and path 190, row 26) from 24 May 2001. Two sets black-and-white aerial photographs from 1956 and 1990, and a set of color orthophotographs from 2002 were obtained from the Litovelske Pomoravi Protected
Landscape Area Administration to serve as ground truth data for the map classification. Vector data of the Czech Republic boundary and the Litovelske Pomoravi PLA area were obtained from the CENIA (Czech Environmental Information Agency) ArcIMS server
[http://geoportal.cenia.cz].
Scene Path/row Acquisition Sensor Format Spatial Bands number date resolution (m) 044-131 204/25 1976-05-08 Landsat MSS GEOTIFF 57 x 57 1 - 4 036-343 190/25 2001-05-24 Landsat ETM+ GEOTIFF 30 x 30 1 – 5, 7 036-344 190/26 2001-05-24 Landsat ETM+ GEOTIFF 30 x 30 1 – 5, 7
Table 1: Acquired satellite images
2.3 Image Processing and Classification
Acquired data sets were processed and examined in the Clark lab’s GIS software
IDRISI 15.0 the Andes edition. Figure 2 shows the steps of image processing and classification needed to achieve defined study objectives. After downloaded from the Global Land Cover
Change Facility and imported to IDRISI, the satellite data were assessed for image quality. While both ETM+ images did not exhibit any significant radiometric noise, the MSS image contained haziness and subtle striping throughout the scene. The Principal Component Analysis was run, using standardized variance/covariance matrix and all four MSS bands as inputs. PCA created four principal component images where the first two explained over 98% of the total variance and
4 the remaining two components contained most of the noise. The original 4 MSS bands were restored through the inverse PCA technique using the first two components.
Cloud-free Landsat Landsat Noise Landsat MSS MSS removal MSS 1976-05-08 1976 (PCA) 4 bands 1976
Global Land Landsat Import Landsat Cover ETM+ Download to ETM+ Change 2001-05-24 IDRISI 2001 Facility 6 bands Mosaiced Landsat Mosaic ETM+ 2001 Landsat ETM+ Landsat 2001-05-24 ETM+ 6 bands 2001
Clipped Resampled Landsat Resample Landsat MSS to 30x30 m MSS 1976 1976 Land cover change Window
Land cover map 1976 Clipped Results Maximum Land Landsat Land cover likelihood Change ETM+ persistence classification Modeler 2001 Land cover map 2001
Land cover trends
Figure 2: Work-flow diagram
The study area is located in the overlap of the two ETM scenes from 2001. A composite of the overlapping images was created using mosaic technique by spatially orienting them and optionally balancing the numeric characteristics of the image set based on the overlapping areas.
In addition, the WINDOW module, extracting sub-images from the set of original images, was utilized to isolate the desired extent of the study area. Finally, spatial resolutions of the images from both times had to be synchronized before the actual image classification. The original resolution of the MSS image (57 x 57 meters) was resampled using parameters from the ETM image (30 x 30 meters), producing a total root mean square error of 0.8 m, which is less than 0.5
5 pixels.
The MSS 1976 and ETM 2001 images were classified using the Maximum Likelihood supervised classification because most of the land cover mapping projects have applied parametric classification algorithms to identify spectrally distinct groups of pixels (Smits et al.
1999). Seven land cover categories were recognized in the Olomouc region: water, deciduous forest, coniferous forest, mixed forest, developed (urban) areas, areas of (intensive) agriculture, and meadows (grassland). Training sites were digitized based on the ground truth data of aerial photographs and orthophotographs. Spectral signatures of urban areas and agricultural areas with bare soil were mixing, therefore their training sites had to be redefined, and also the texture analysis using Dominance index and kernel window of 5x5 pixels was conducted. Finally, the
Maximum Likelihood classification was run with original bands and the texture image as inputs, producing two final land cover maps of 1976 and 2001 that were compared.
A crossclassification procedure is a fundamental pairwise comparison techniques used to compare two images of qualitative data (Eastman 1995). IDRISI Andes offers efficient tools for rapid assessment of land cover change and its implications based on crossclassification principles. The Land Change Modeler (LCM) for Ecological Sustainability allowed using the classified land cover maps from 1976 and 2001 as input parameters and identifying the locations and magnitude of the major land change, land persistence, and transitions between land cover categories in the study area.
3. RESULTS
Figures 3 and 4 represent the results of Maximum Likelihood classification: land cover maps from 1976 and 2001. The change analysis tool provides statistical assessment of changes in
6 individual land cover categories. Its result in figures 5 and 6 demonstrates that there have been significant changes in all land cover/land use categories between 1976 and 2001 with the exception of water, where the subtle change can be caused by location errors in land cover classification. Concerning the net change, which represents the earlier area of a category with added gains and subtracted losses, three land cover categories experienced major transitions. The total area of meadows (grassland) increased by 942 km2, while the area of intensive agriculture decreased by 592 km2, as well as the area of coniferous forests which decreased by 603 km2. Also the category of developed (urban) was affected by distinct change with net gain of 127 km2.
o
Figure 3: Land cover map of 1976
7 o
Figure 4: Land cover map of 2001
Figure 5: Gains and losses between 1976 and 2001 in km2
Figure 6: Net change between 1976 and 2001 in km2
8
A simplified crossclassification map (figure 7) represents persistence in land cover categories, areas where no change occurred, and land cover change, areas with any type of transitions between categories (depicted in black). However, the land change and persistence map is difficult to interpret visually. Therefore, the contribution to net change, i.e. the transition between specific classes, was calculated. Data in figure 8 represent the contribution to net change for categories of meadows (grassland), developed (urban), and mixed forest. They reveal that agricultural areas explain about 63% of the total increase in meadows, new development occurred predominantly on former agricultural areas (over 56%), and about 16% of previous coniferous forests is currently identified as mixed forest.
o
Figure 7: Land cover persistence and change
9
Fig. 8: Contribution to net change in selected categories (km2)
The trend analysis helps to decipher the location of complex land cover changes by providing a mean of generalization about the transition between selected categories. Three maps of trends (figure 9) were created depicting the transitions from 1976 to 2001 between categories of interest: agricultural to meadows, agricultural to developed, and coniferous forest to mixed forest. The resulting maps represent simulated surfaces where the values have no special significance but denote the generalized locations of transitions from areas with no change to areas with significant change.
no change
Agricultural to meadows
change
o
10
Agricultural to developed no change
change
o o
no change Coniferous to mixed
change
o
Fig. 9: Trends in transitions between selected categories
4. DISCUSSION AND CONCLUSIONS
The results support initial assumptions based on general knowledge of some of the land use drivers in different times. There have been significant losses in categories of intensive agricultural areas and coniferous forest, and gains in meadows and developed areas. From the former agricultural areas, 23% became meadows and pastures, especially in the northeastern hilly part of the study site, and 3% were developed in the lowlands around the Litovelske Pomoravi
11 Protected Landscape Area. About 16% from the previous coniferous forest in the eastern hilly part of the region was currently identified as mixed forest.
This study provides no empirical evidence of direct causality between discovered land cover changes in the study area and political transformation of the country. However, observed marginalization of agricultural areas is consistent with suggestions of Zemek et al. (2005) that it occurs at locations with unfavorable natural conditions, especially in uplands. Similarly, new developed areas were found especially in central lowlands near existing cities which is consistent with general suburbanization process in central Europe (Ptáček 1998). Finally, transitions from coniferous tree cover to mixed forest was observed in the northeastern part of the study site, which correlates with general diversion in forest management from spruce plantations to an alternative use of native species of trees.
Classification of multispectral satellite data and comparison of land cover maps is an essential tool for assessing large-scale land cover/land use changes. However, the research left a plenty of open space for future improvement. For example, there was no empirical validation of classification accuracy involved in the analytical process. Spectral mixing was apparent between classes of developed and agricultural, phenological stages of crops in the time of data acquisition exhibited similar spectral response as meadows, or the MSS imagery suffered from radiometric noise. An effort to collect better quality remote sensing data should be made to improve overall accuracy of land change assessment. Finally, an alternative to Maximum Likelihood classification, such as Decision Tree classification techniques, should be considered for future improvement of the assessment accuracy, as it learns probabilities of land cover classes from the distribution in the training data (Rogan et al. 2002).
12 5. ACKNOWLEDGMENTS
I would like to thank Yelena Ogneva-Himmelberger and John Rogan, professors from Clark
University, for supervising my work and the Fulbright program for enabling my studies in the
United States.
6. ABOUT THE AUTHOR
Tomas Vaclavik was a Fulbright exchange student in the program of Geographic Information
Sciences for Development and Environment at Clark University, Worcester, MA, USA in the academic year 2006/2007. He earned both his Bachelor’s and Master’s degrees in Ecology and
Environmental Sciences at Palacky University at his home country Czech Republic. He has been recently working as a GIS analyst for the Agency for Nature Conservation and Landscape
Protection of the Czech Republic. Currently, he pursues his PhD in Geography at University of
North Carolina at Charlotte, focusing on applications of GIS in ecological research.
13 7. REFERENCES
Eastman, J. R. (1995). Change and Time Series Analysis. United Nations Institute for Training
and Research/Clark Labs for Cartographic Technology and Geographic Analysis, Clark
University. Worcester, MA 01610, USA: 28.
Fanta, J., Prach, K. and Zemek, F. (2004). Status of Marginalisation in Czech Republic:
Agriculture and Land Use. EUROLAN, National Report – Czech Republic. České
Budějovice, Faculty of Biological Sciences, University of South Bohemia, and Institute of
Landscape Ecology - Academy of Sciences of the Czech Republic: 29.
Fanta, J., et al. (2005). Strengthening the multifunctional use of European land: Coping with
marginalization. EUROLAN, National Report – Czech Republic. České Budějovice,
Faculty of Biological Sciences, University of České Budějovice, Institute of System
Biology and Ecology, Academy of Sciences of the Czech Republic: 41.
Jackson, J. (2002). Urban Sprawl. Urbanismus a územní rozvoj 5(6): 21-28.
Neuhäuslová, Z., et al. (1998). Map of Potential Natural Vegetation of the Czech Republic.
Prague, Academia.
Ptáček, P. (1998). Suburbanizace - měnící se tvář zázemí velkoměst. Geografické rozhledy 7(5):
134-137.
14 Ptáček, P. (2000). Networking and Local Culture: Local CommunityTransformation after 1989
on the Example of Olomouc, Czech Republic. Acta Universitatis Palackianae –
Geographica 36: 59-64.
Rogan, J., Franklin, J. and Roberts, D. A. (2002). A comparison of methods for monitoring
multitemporal vegetation change using Thematic Mapper imagery Remote Sensing of
Environment 80(1): 143-156.
Smits, P. C., Dellepiane, S. G. and Schowengerdt, R. A. (1999). Quality assessment of image
classification algorithms for land-cover mapping: a review and proposal for a cost-based
approach. International Journal of Remote Sensing 20(8): 1461-1486.
Zemek, F., et al. (2005). Multifunctional land use – a chance of resettling abandoned landscapes?
(A case study of the Zhůří territory, the Czech Republic). Ecology 24(1): 96-108.
15 NOTE
This is a condensed version of the original paper submitted to the URISA Student Paper
Competition 2007. The author is currently extending his research, using different data, new methods and validation techniques, and intends to publish his results in a professional remote sensing journal. This condensed version has been created in order to avoid an overlap with the intended future publication.
16