INTERNATIONAL SCIENTIFIC CONFERENCE AND XXIV MEETING OF SERBIAN SURVEYORS ″PROFESSIONAL PRACTICE AND EDUCATION IN GEODESY AND RELATED FIELDS″ 24-26, June 2011, Kladovo - ,,Djerdap“ upon Danube, Serbia.
DASYMETRIC MAPPING OF SPATIAL DISTRIBUTION OF POPULATION IN TIMOK REGION Branislav Bajat1, Nikola Krunić2, Milan Kilibarda1 1
University of Belgrade, Faculty of Civil Engineering, Department of Geodesy and Geoinformatics, Belgrade, SERBIA, E-mail:
[email protected],
[email protected] 2 Institute of Architecture and Urban&Spatial Planning of Serbia, Belgrade, SERBIA, E-mail:
[email protected] Summary: Dasymetric mapping of population distribution represents very functional visualization method used in spatial demographic analysis. The main advantage of dasymetric mapping over standardized cartographic methods (choropleth maps) used for population density mapping is its ability to realistically place population data over predefined geographic space. The application of dasymetric population mapping becomes wider with the expansion of powerful and efficient GIS tools and accessible public domain spatial data bases on the WEB. In the work presented in this paper, spatial distribution of population was modeled by using three class model of dasymetric mapping. Census data of the year 2002 were deaggregated by utilization of CORINE 2000 land cover data as ancillary predictors. The eastern part of Republic of Serbia, actually Timok Region was used as a case study area in this research. Keywords: dasymetric mapping, population distribution, CORINE 2000, Timok Region
1. DASYMETRIC MAPPING The use of maps in demographic analyses is inevitable when assessing spatial aspects of demographic changes and movements and their interactions with the environment. Data on population are processed based on the census conducted by states at regular time intervals, most often once in a decade. In the Republic of Serbia, the data are normally presented on the level of census designation places (by aggregation of census and statistical cycle data), in the USA those are census blocks, and in Great Britain enumeration districts. Demographic census data are mapped as statistical surfaces [3] and most often presented on choropleth maps [5]. The model offered by a choropleth map is the result of aggregation of data obtained in census cycles. The data such as population density shall in that case result in surfaces which do not envisage presentation of uninhabited areas, although these actually exist. The only way to overcome this problem, towards the objective of the most realistic possible modeling of demographic data, is utilization of spatial bases which indicate to the lot coverage and spatial contents. Dasymetric mapping method is one of the possible approaches to solving of this problem, dividing the modeled space into zones of higher homogeneity degree, thus reflecting more truthfully the variations in a statistical layer, with support of additional variables and their correlations. Mennis [11] defines dasymetric mapping as a process of distribution (deaggregation) of spatial data per smaller spatial units, more suitable for analysis, using additional/auxiliary data, in order to produce a finer distribution of population or other spatial phenomena. Application of this method dates back to the 19th century, and the first cartographer who used this technique was George Julius Poulett Scrope in 1833, mapping the classes of global population density. The Russian geographer Tian-Shansky, who described the method in 1911 and whose map of European Russia population distribution was published in 1923 [1], is most frequently referred to as the first author of a dasymetric map. However, John Kirtland Wright was the first one to introduce the method and the origin of the word “dasymetry” in English language, in 1936, explaining it as “density measuring”. Although the choropleth mapping was practiced an
1
entire century before Wright, the neologism ‘choropleth’, meaning ‘value-by-area’ method, is still attributed to him [8], [10]. In the course of time, the techniques of dasymetric mapping have been improved and multiplied by means of various data deaggregation or interpolation methods. Outline of methods used in dasymetric mapping may be found in the paper written by a group of authors, [10]. Although the ultimate goal of dasymetry is generation of a map which depicts the spatial distribution of population as truthfully as possible, the methodology applied towards its acheivement is based on diverse concepts [10]. In this paper, the “three classes” method is applied. The three classes method, using the categories of land use, attributes the percentage of a class’s participation in total number of inhabitants in particular area (a well-known example is that the class of urban land is attributed e.g. 70%, agricultural/forest/non-urban land 20%, whereas 10% goes to forest land [4]. The percentage of participation is certainly determined on case-to-case basis and depending on analyst's arbitration. Obviously, this method still „suffers“ from the issue of delineation between urban and other land, and homogenizes density within classes. From geodetic aspect, the cadastre-based expert dasymetric system is found interesting (cadastral-based expert dasymetric system - CEDS). Maantay et al. [10], used cadastral data as predictors in large-scale dasymetric mapping. The expert system was used to identify the cadastral data, as an auxiliary variable, would best depicts the actual number of population. The modeled population always preserves the pycnophylactic property, which was rarely achieved in previous methods. According to these authors, the major advantage of CEDS method is high precision of population distribution in high resolution in densely developed urban blocks with heterogeneous population. The technique implies utilization of data on residential zones and housing units, supported with taxation data, per lots. The selection between a residential area and housing unit is made by an expert system. The results were checked in comparison with census data and other dasymetric techniques, in order to improve the method. In our practice, the dasymetric mapping method was successfully applied on entire region of the Province of Vojvodina, on the example of “daily” and “night” population mapping, towards the most truthful possible presentation of daily fluctuations and migrations of working population [9].
2. CASE STUDY AREA OF TIMOČKA KRAJINA The data referring to geographic-demographic characteristics of surveyed area were collected for the requirements of Regional Spatial Plan of Timočka Krajina [6]. 2.1. Position and Main Features of the Region The region covered by the Spatial Plan, over total area of 7,130 km 2 (about 8% territory of the Republic of Serbia), occupies the eastern part of the Republic of Serbia and covers the territories of the Zaječar and Bor administrative districts. In physical-geographic terms, it covers most of the Timok basin, part of lower Podunavlje and the zone of its hilly-mountainous hinterlands, the upper - spring zone of the Pek river watershed and upper and mid part of the Sokobanjska Moravica watershed. The Timočka Krajina region is surrounded by the Republic of Romania in the north, the Republic of Bulgaria in the east, the Niš and Pirot administrative district in the south, and Braničevo and Pomoravlje administrative district in the west. The Spatial Plan region belongs to underdeveloped and both economically and demographically depressive, specific-purpose regions: the region of the Pan-European transport corridor VII „Danube“ and contact area between Pan-European infrastructure corridors X in the west and IV in the east; the region with outstanding hydropower potentials (Hydropower and Navigation System Djerdap); an agricultural-cattle breeding and forest region; natural and tourist attractions (development of tourism on the Stara Planina mountain and on the Danube); water springs ranked national or regional ones, etc.; a region with significant reserves of mineral resources and developed mining industry (Mining and Smelting Complex „Bor“), etc. 2.2. Population and Settlements The population of the Spatial Plan region numbers around 284,100 inhabitants, i.e. 3.8% total population of the Republic [13], with constantly declining tendency in all census periods since 1948. Average population density is 40 p/km2, which is less than a half of the republic average (85 p/km 2). The population is relatively evenly distributed per administrative districts of Timočka Krajina. The settlement network of the Region is a system of 263 settlements situated in 267 cadastral municipalities. The total of 11 settlements have the status of urban ones, with population of about 152,750 (53.8% of the Spatial Plan region). Average size of a settlement area approximates 26 km 2. Differences among districts are quite pronounced, both in terms of settlements, cadastral communes and size of areas, and the function of centers. In the Bor district, there are 98 cadastral municipalities, with 90 settlements and average size of areas around 35.8
2
km2, whereas in the Zaječar district there are 169 cadastral municipalities, 173 settlements, with average size of areas around 21.4 km2. Bor plays several roles in spatial-functional organization of the Republic of Serbia, administrative district of Bor and municipal area, in fact, it represents the functional center, with 13 municipal settlements and indirectly 76 settlements of functional region gravitating towards it. Besides, Bor is the center of regional urbanization (app. 56,000 inhabitants, i.e. about 71% population of the municipality) with functionalintegration processes which exceeded the scope defined by territorial and administrative organization of the Republic of Serbia. The role of Zaječar in spatial-functional organization of the Republic of Serbia, administrative district of Zaječar and town territory, is reflected in the following: it is a functional center with 41 settlements of the municipality and indirectly 132 settlements of the functional region gravitating towards it; some of the settlements from the Bor functional region also gravitate towards it (settlements in the southern part of the Negotin municipality); about 60% town population or about 28% district population is concentrated in it; it is the center of regional urbanization of Timočka Krajina and of the Timok development belt (Timok development axis) which spatially-functionally integrates the east Pomoravlje with the Pirot and Nišava administrative district and Veliko Pomoravlje; in the domain of functional-integration processes, it exceeds the scope defined by territorial and administrative organization of the Republic of Serbia. [7].
3. DASYMETRIC MAPPING OF TIMOČKA KRAJINA POPULATION Dasymetric modeling generated a map of spatial distribution of Timočka Krajina population. For this occasion, a very simple dasymetric model was used, with land cover data of CORINE 2000 [12] as basic predictor. Percentage values of population participation per classes of land cover are based on certain experience and papers so far [4]. The paper made use of standard ArcGIS environment for generation of visual presentations and dasymetric model. The standard choropleth map based on the last census data from year 2002 [13] is shown in Figure 1. (left); mapping unites are polygons that depict census designation places. Dasymetric map of the same region (Figure 1. right) indicates a more realistic presentation of spatial population distribution. The differences in those two maps are apparent; the choroplet map reflects dummy results, which are especially marked for class that indicates density between 51-100 p/km 2 (shown with red color). Actually those areas are unpopulated spaces with sparsely distributed settlements.
Figure 1: Population density maps of Timočka Krajina; choroplet map (left) and dasymetric map (right). In order to get better insight of the results obtained, large scale representation for particular towns, like Bor and Zaječar, is shown in Figure 2. Attention is drawn to differences in modelling the population density between
3
city/municipal centers and surrounding settlements, as well as to the differences occurring within the city/municipal centre itself (between industrial and sport-recreation surfaces, and other city tissues).
Figure 2: Population density of Bor and Zaječar towns; choroplet map (left) and dasymetric map (right). Both maps are in the same scale. The objective of further research and improvement of this method in national theory and practice shall be the adjustment of the initial model, in order to eliminate, or reduce to the minimum, the subjectivity and tentativeness in the process of modeling. The usage of additional data bases, like network of settlements digitized from topographic maps TK100 or TK50, or WEB accessible public domain spatial data of soil sealing [2] should be of great benefit in reliable population distribution modeling.
4. FINAL CONSIDERATIONS The ever more accessible GIS packages intended for modeling of spatial data open up new possibilities in processing and visualization of demographic data. The option of combining demographic and other spatial databases, such as CORINE data base, is particularly important. The possibility to apply dasymetric modeled spatial distribution of population is vast and important, primarily for analyses and projections in spatial and urban planning, accident risk assessment, hazard control, environmental protection, socio-economic disciplines, etc. It is reasonable to expect increased interest of scientific and expert public in application of this method, thus considerably advancing the methods and models and, the most important of all, reaching a higher level of research coverage of the space, phenomena and processes in the Republic of Serbia.
ACKNOWLEDGMENT The paper represents the result of research carried out on projects No. III 47014 „The role and implementation of the National spatial plan and regional development in renewal of strategic research, thinking and governance in Serbia“ and TR 36035 „Spatial, ecological, energy and social aspects of settlements’ development and climate changes - interrelationships“ financed by the Ministry of Education and Science of the Republic of Serbia.
REFERENCES [1] Bielecka E.: A dasymetric population density map of Poland. (2004). Available on Web site: http://www.cartesia.org/geodoc/icc2005/pdf/oral/TEMA5/Session%209/ELZBIETA%20BIELECKA.pdf
4
[2] Burghardt W.: Soil sealing and soil properties related to sealing. Geological Society, London, Special Publications 266, (2006), p. 117-124. [3] DeMers M. N.: Fundamentals of Geographic Information Systems. (2nd ed.) Published by John Wiley, New York, (1999), pp. 498. [4] Eicher, C. L., and C.A. Brewer.: Dasymetric mapping and areal interpolation: Implementation and evaluation. Cartography and Geographic Information Science 28, (2001), pp. 125-138. [5] Harvey F.: A primer of GIS: fundamental geographic and cartographic concepts. Published by The Guilford Press, New York, (2008), pp. 310. [6] IAUS: Regionalni prostorni plan Timočke krajine, Nacrt plana, Institut za arhitekturu i urbanizam Srbije i Republička agencija za prostorno planiranje. Beograd. (2010) Available on Web site: http://www.rapp.gov.rs/index.php?kuda=dummy&sta=planovi&idplana=125 [7] IAUS: Regionalni prostorni plan Timočke krajine, Dokumentaciona osnova, ekspertiza „Razvoj mreže naselja“. Published by Institut za arhitekturu i urbanizam Srbije, Beograd (2010). [8] Jarcho S. M. D.: Some early demographic maps, Bull. N. Y. Acad. Med. Vol. 49(9), (1973), pp. 837-844. [9] Krunić N., Bajat B., Kilibarda M., Tošić D.: Modelling the Spatial Distribution of Vojvodina’s Population by Using Dasymetric Method. Spatium 24, (2011), pp. 45-50. [10] Maantay J. A, Maroko A. R, Herrmann C.: Mapping population distribution in the urban environment: the cadastral-based expert dasymetric system (CEDS). Cartography & Geographic Information Science, 34 (2), (2007), pp. 77-102. [11] Mennis J.: Generating surface models of population using dasymetric mapping. The Professional Geographer, Vol. 55 (1), (2003), pp. 31–42. [12] Nestorov I, Protić, D.: CORINE kartiranje zemljišnog pokrivača u Srbiji. Published by Građevinska knjiga, Beograd, (2009), pp.180. [13] Statistical Office of the Republic of Serbia: 2002 Census of Populations, Households and Dwellings (in Serbian). (2003) Belgrade.
5