Water in most important towns of the Czech Republic

ABSTRACT

An analysis of surface water presence of one hundred most important towns of the Czech Republic has been conducted with respect to their proportion to town area. The results of the analysis, which was performed on the basis of data from specialised water management database, were visualised in a poster clearly showing ‘blue areas’ in given towns. In relation to the visualisation purpose, which was the comparison of towns, the towns were presented so that their area was identical. The order of towns in the poster was established in accordance with the representation of blue areas. As supplementary information, the total area of blue areas per capita was displayed in visual form. The poster is accompanied by a map showing the location of the towns within the Czech Republic.

KEYWORDS:

• Water courses
• water bodies
• blue areas
• water in town
• Czech Republic

1. Introduction

Being a condition for life as well as an important landscape element, water has always been the central concern of human society. There has been and still is a significant close relation between settlements and water courses and/or water bodies. These relations have been the content of numerous geographic studies such as that of Paul and Meyer (2001). The relations have varied through time, which is presented by Heikkila (2011), clearly defining pre-industrial, industrial, and post-industrial periods.

Since relations between a town and water (water course, water body) are complex, they are often studied from different perspectives, e.g.:

• The influence of a town on a river in terms of: (i) discharge and water quality, as studied by Zaharia, Ioana-Toroimac, Cocoş, Ghiţă, and Mailat (2016), (ii) changes in morphology, as studied by Chin (2006), (iii) the content of nutrients in water, as studied by Meyer, Paul, and Taulbee (2005), and many other similar approaches;

• The influence of a river on a town that is most often studied with respect to flood hazards. However, currently there is an intensive study of other aspects of the influence of a river on a town, especially what a river provides a town with in a form of ecosystem services (a review of water valuation literature is presented by Doherty, Murphy, Hynes, & Buckley, 2014). Even though less studied than terrestrial ecosystems (Ziter, 2016), urban aquatic ecosystems significantly affect human life. Besides commonly evaluated regulatory services such as climate affecting factors or flood hazard mitigation (Villarreal, Semadeni Davies, & Bengtsson, 2004), water in town allows for social contacts (boat transport) or, depending on the river width, divides society (Kondolf & Pinto, 2017). Also, waters stand for an important source of services related to recreation, health, aesthetics and education – Elmqvist et al. (2015) presented them as ‘non-monetary benefits'. In this respect, water bodies are often studied along with urban forests and parks and they are referred to as ‘green and blue infrastructure' (Elmqvist et al., 2015), ‘green and blue space' (Sirakayaa, Cliqueta, & Harrisb, 2018) or ‘green and blue areas' (Bolund & Hunhammar, 1999). Studies dealing with ecosystem services are often related to big cities. However, a few studies exist concerning blue areas in smaller towns (according to the population), as for the territory of the Czech Republic, e.g. in those of Jakubínský and Báčová (2013).

Waters are one of the basic elements of map content and therefore commonly displayed in maps. In the case of thematic maps presenting waters in towns, real or modelled overflow, which in many cases makes a part of land planning documents, is most often displayed (e.g. Di Salvo, Ciotoli, Pennica, & Cavinato, 2017; Gonçalves, Inês Marafuz, & Gomes, 2015). On the other hand, sporadic maps are those presenting groundwater in the territory of a town such as that of La Vigna et al. (2016), or water consumption maps such as ‘The City of Cape Town’s water map' (https://citymaps.capetown.gov.za/waterviewer). Maps make not only the resulting product, but also an important source of information. For example, Visser (2014) analyses historical maps to identify the causes of floods.

For the purposes of this study, the term ‘blue areas’ was borrowed from the above-cited authors (Bolund & Hunhammar, 1999; Elmqvist et al., 2015; Sirakayaa et al., 2018); it refers to the open surfaces of (flowing) watercourses and (standing) bodies of water, i.e. streams, rivers, lakes, reservoirs and ponds. ‘Blue areas’ does not cover all water surfaces, because marshlands may also be classified as water surfaces (e.g. in the CLC classification – Kosztra & Arnold, 2014), and in this study marshlands are not included in the ‘blue areas’ category. Areas of seawater may also be included within the territory of a city (e.g. coastal lagoons), but this does not apply to the Czech Republic.

The concise, easy-to-understand and already-established term ‘blue areas’ – based on the association of water with the colour blue – makes it possible to avoid the repeated use of the relatively lengthy description ‘watercourses and bodies of water’. If the term ‘bodies of water’ were used, the designation could potentially be erroneously assumed to refer to lakes, reservoirs and ponds only, thus excluding watercourses. The term ‘blue areas’ is currently used exclusively in the evaluation of geo-ecological services, but it could also potentially be applied in other fields such as spatial planning, landscape protection, tourism and recreation, etc. – in other words, in all fields where watercourses and bodies of water are grouped together into a single category. However, the use of the term is not appropriate when it is necessary to retain a distinction between watercourses and bodies of water.

2. Aims

The aim of this study is to undertake an evaluation of the occurrence of water surfaces in urban areas using a relatively unusual method. Many existing studies offer detailed assessments of water surfaces in urban areas from a variety of perspectives (e.g. Doherty et al., 2014 lists 15 variables that can be evaluated). Such detailed approaches are used when evaluating individual cities or comparing two cities. The goal of this study was to choose just one parameter and to evaluate a large number of cities and towns based on this single parameter. The chosen parameter represents a basic means of characterising the occurrence of water surfaces in cities – the occurrence of blue areas expressed in proportional terms. The study focused primarily on the proportion of blue areas out of the total area of the town (in %) and the total area of the blue areas per capita (in m2). The comparison applied to the top hundred cities and towns in the Czech Republic.Undertaking the study, the following three aims were being pursued:

1. Based on available suitable data, the identification of blue areas in the most important towns of the Czech Republic;

2. Assessment of obtained data: in particular the mutual comparison of the towns according to the representation of waters on town area and the comparison according to the total area of the blue area per capita;

3. Adequate visualisation of the results in a form of a poster.

3. Selection of towns

Main criteria for the selection of the towns are territorial management and population. The Czech Republic is composed of 77 districts that are joined in 14 higher autonomous units – regions, while 14 regional capitals are at the same time district municipalities. The fifteenth region is the capital city of Prague that lacks the statute of a district municipality. For the purpose of the study, a total of 100 towns are selected in accordance with the following scheme:

• the capital city of Prague,

• 77 district municipalities (14 out of which are regional capitals),

• another 21 towns according to the population number,

• Vodňany town – the reason for this choice is the name of the town that directly refers to the topic of the study; the Czech name of the town can be translated into English as ‘Water Town'. As Hristova, Dragićević Šešić, and Duxbury (2015) state: ‘In history, its “water magic” was mentioned in Medieval Latin texts as Aquilea Bohemorum’. The population of 6,856 inhabitants does not range Vodňany among the first hundred most populated towns, but there is not a big difference between this town and the smallest town in the list (Semily on the 99th position with 8,447 inhabitants).

The total population of the selected towns is 5.1 million inhabitants, which is nearly half the population of the Czech Republic (10.5 mil. inhabitants).

4. Data sources and processing

The values of the representation of blue areas in the towns of the Czech Republic were not any official data. No freely available summary data are available on the occurrence of watercourses and bodies of water in individual cities and towns in the Czech Republic. Therefore, to obtain the information, it was necessary to carry out a joint analysis comprising both the analysis of hydrological data and the data on towns.

4.1. Hydrological data selection and used data characteristics

To meet the objectives, three types of hydrological data were available:

1. CORINE Land Cover (CLC) data covering the whole territory of the Czech Republic. CLC is a European Union project coordinated by the European Environment Agency. CLC uses data from the LANDSAT satellite images to map land cover in Europe at a scale of 1:100,000. Additional information is taken from aerial photographs, maps and field research data. Land cover is described using a nomenclature consisting of 44 categories organised hierarchically in three levels; each category is assigned to one of five main CLC land cover classes. The five main classes – which comprise Level 1 (the highest level) in the hierarchy – are as follows: artificial surfaces, agricultural areas, forests and semi-natural areas, wetlands, and water bodies. Inland waters consist of two categories: water courses and water bodies. A disadvantage of using CLC data is low data accuracy. Water courses are only recorded if being at least 100 m wide, smaller ones are missing. In the description of the category of water bodies, the smallest vectorised water body fails to be indicated, therefore the basic size of a minimal mapped polygon (25 ha) is applied. For the lack of detail, water bodies and water courses from the CLC geodatabase were not used.

2. Cadastre data in which water bodies and water courses belong to basic categories. The advantage of these data is that they are very detailed – cadastre maps in the Czech Republic are managed at a scale of 1:2000. A disadvantage of these data is disparity in the processing method within the territory of the Czech Republic and the way of the specification of water bodies that are defined by land ownership boundaries, not the real waterbody boundaries. For this reason, water bodies and water courses from the cadastre data were not used.

3. Sets of data coming from the ZABAGED database. ZABAGED (The geographic base data of the Czech Republic) is a complex digital geographic model of the Czech Republic at a scale of 1:10,000 that contains spatial and descriptive information on residences, communications, waters, territorial units, protected areas, vegetation etc. The database, which is administered by the Land Survey Office, is used to create the state map series. Although Taylor and Stokes (2005) consider using a ‘blue line on a topographic map’ to specify water course boundaries as merely schematic, ZABAGED data were collected by means of direct mapping. They serve as a source for a range of products used by the state and other authorities. Data sets of ArcČR 500 and DIBAVOD used in this study have also been derived from this database.

The polygons of selected towns were obtained from the layer of municipalities of the Czech Republic (CR) from the ArcČR 500 digital vector geographic database of the CR. The ArcČR 500 database is at a scale of 1:500,000 and was created as a joint venture by three partners: the company ARCDATA PRAHA, s. r. o.; the Land Survey Office; and the Czech Statistical Office. The database is freely available on the ARCDATA PRAHA website (https://www.arcdata.cz/). The general terms of the licence state: The user is entitled to use the Data for the purposes of creating cartographic works as part of projects implemented for third parties. The user is obliged to acknowledge the source of the Data on the cartographic work using the following format: ‘ArcČR, ARCDATA PRAHA, ZÚ, ČSÚ, 2016’ [ZÚ = Land Survey Office, ČSÚ = Czech Statistical Office] (https://www.arcdata.cz/).

To analyse the occurrence of water bodies and water courses, we used freely available data from DIBAVOD (a digital hydrological database), which is a project of T. G. Masaryk Water Research Institute. DIBAVOD is a reference geographic database created from corresponding ZABAGED layers and intended for the creation of thematic cartographic outputs with the theme of water management (Košut & Levitus, 2017). The catalogue of DIBAVOD objects, which is continuously updated, contains 75 objects in 10 specific categories. Out of DIBAVOD data, we used the following layers from the category of ‘underlying phenomena of surface waters and groundwater’:

• polygon layer of water bodies – the data contain both anthropogenically established reservoirs (including ponds) and natural water bodies (lakes),

• watercourse boundary of water courses representing streams wider than 5 m,

• line layers of water courses representing the centrelines of water courses having width smaller than 5 m.

All data are current to the date of imaging, which takes place every two years for the whole territory of the CR.

4.2. Calculation of town areas and blue areas

The boundaries of towns were taken from the above-mentioned ArcČR 500 database in the form of polygons on the basis of which the areas were calculated.

Polygons of water bodies were used directly for the calculation of area. Watercourse boundaries of wider water courses were first converted into polygons.

In the case of the water courses of the width smaller than 5 m, the lengths were known, but no more precise information about the widths are available. Theoretical width of these water courses is 0–5 m; the mean width is thus 2.5 m. Despite the fact that the information concerning the ZABAGED database contains no information on the minimum width of water courses in this database, there is a high probability that water courses of the width from 0 to 1 m were not mapped at all or just minimally (1 m at a scale 1:10,000 represents water course width of 0.1 mm). On the basis of this reasoning, the estimated mean width was shifted from 2.5–3.0 m. For the water courses of the width less than 5 m, the mean width was then supposed to be 3 m. The lengths of water courses were then multiplied by this value in order to obtain the estimate of their area.

Calculated areas were influenced by cartographic distortion and therefore they were adjusted according to the method described in Chapter 4.3.

4.3. Cartographic representation

Used ArcČR 500 and DIBAVOD datasets are in the Křovák’s projection in the national coordinate JTSK system EastNorth. Kennedy and Kopp (2000) characterise this projection in detail under the name ‘Krovak’; according to EPSG it is marked as ‘5514 (S-JTSK/Krovak East North)’, (https://epsg.io/5514). Křovák’s projection is conformal conic projection in oblique location. It is official projection used in the Czech Republic for all civilian purposes including the real estate cadastre and state mapping.

An advantage of conformal projection is that the shape of small territories remains preserved, which is the case of our projected towns. On the other hand, a disadvantage of conformal projection is area distortion, which is the main content of this work. Although Křovák’s projection is designed with respect to the minimalisation of the length distortion, which is maximally 14 cm per 1 km at the extremity of the territory, it influences the distortion of the size of towns and water bodies, depending on the position – to obtain real areas, equal-area projection would have had to be used. To make a common map, preserving both conformal and equal-area projection is impossible, but the selected method used to visualise the towns independently made it possible to use advantages of both the types of projection. For each town, the values of length distortion (see Table 1) and areal distortion were calculated that were subsequently used to correct the lengths and areas obtained from the conformal projection. In this way, correct values of the surface area of towns and blue areas (presented in Table 1) were acquired.

Table 1. Input data and results of analysis of blue areas in most important towns of the Czech Republic.

Note: Blue columns – data from DIBAVOD database; Red columns – data from ArcČR 500 database; Green columns – values corrected for distortion.

Calculations of values in columns: K = C. 1000. 3/J, L = (D + E)/(J.J), M = K + L, N = G/(J . J), O = M . 100/N, P = M/F.

^aFull name: Brandýs nad Labem-Stará Boleslav.

To visualise the towns, data in conformal projection were applied, i.e. the shapes were preserved. For each town, the scale was calculated so that the area of towns represented on a poster was identical and when calculating the scales, we started from the area of corrected distortion.

In order to produce a overview map, the original Křovák’s projection was left unchanged. Therefore, with regard to the scale (1:1,500,000), the impact of distortion on the area of towns is not evident.

5. Results

Basic results for individual towns are the following: the total area of all blue areas, their percentage in the area of town, and their size per capita. The results are presented in Table 1, along with the input data.

The resulting values show that the highest proportional spatial representation of blue areas is identified in the Cheb town where water covers more than one tenth of the town territory (11.2%). This fact is given by the presence of two water reservoirs whose total area reaches 1000 ha. At the same time, Cheb has absolutely highest sum of blue areas within all studied towns. The second position in the proportional representation of waters is held by the Vodňany town (reaching 7.8%), chosen for its name referring to water, as mentioned above. Three towns follow with the blue area representation ranging from 7.4–7.0%, and 7 towns ranging from 5.4–4.0%. Other towns then reach the value that is smaller than 4%. 21 towns show the representation of blue areas smaller than 1%, the town on the last position is Prachatice with the value of 0.38%. Prague, the capital city and at the same time the largest city in the CR, occurs at the end of the first half of the list, occupying the 44th position as for the proportional representation of blue areas (1.8%). The results can be studied in detail both in the table and in the visualisation as in both the cases the ordering of towns is carried out according to this parameter.

Another result is the size of blue area per capita. This ranges from 414 m² (Vodňany town) to 3.6 m² (Prostějov town). The second position is held by Cheb with 333 m² blue area per capita. The next four towns show the values between 178 and 106 m². The results are presented in Table 1.

5.1. Evaluation of the results and potential applications of the results

For the territory of the CR, there are no aggregated data obtained in accordance with the same or similar methodology such as that used for the calculations in Table 1. The only possibility is to use cadastre data, but with restrictions presented in Chapter 4.1. According to these cadastre data, the spatial representation of blue areas in the CR in 2016 was equal to 2.1%. According to the comprehensive analysis of all towns, the spatial representation of blue areas in towns makes 2.23%, which is slightly above the value for the whole CR, but this favourable value is mainly based on high values at the beginning of the list. Most towns contain ‘less water’ than is the value for the CR as only 36 towns exceed this value and 64 towns are below it.

To compare Czech towns with some towns outside the CR, we can mention Stockholm for which Bolund and Hunhammar (1999) report 13% of blue areas, stressing: ‘This is considerably more water (and green) space than possessed by most other cities, and it gives Stockholm its unique character.’ Geographically nearer is Poznań in which water occupies 3% of its area (Zepp, Mizgajski, Meß, & Zwierzchowska, 2016).

There are significant differences in the distribution of blue areas in Czech towns, the ratio between the largest (11.2%) and smallest (0.38%) area representation is about 30:1. Calculating the same ratio with respect to blue areas per capita, the result is 115:1. These ratios clearly show that the differences between individual towns are considerable. Comparing the towns on the basis of the above-presented parameters, a crucial role is also played, besides absolute size of blue areas, by population density. In this context, it is also worth pointing to towns that are characterised by a small percentage of blue areas, but on per capita basis they moved up tens of positions in the list. Such ‘jumpers’ are primarily Jeseník, Trutnov and Klatovy that moved from the positions at the very end of the list upwards by circa 40 positions.

In addition to the parameters indicated in Table 1 and displayed in the map, what was specified within input data was the relationship of dependency between the area of town on one side and the area of water courses and water bodies on the other side. As for the relationship between the area of town and the area of water courses, the Pearson correlation coefficient was 0.85. This value points to a potential linear relation between these quantities. In the Czech Republic, there are not great differences with respect to river network density (there are e.g. no arid areas), therefore the result corresponds to a logical assumption: larger area – more rivers. On the other hand, the relation between the town area and the area of water bodies (lakes, reservoirs, ponds) shows no linear dependency, the correlation coefficient is 0.39. Therefore, the relation larger area – more water bodies does not apply here.

We also determined the relation between the size of blue areas per capita and the population density in the town. Here it would be possible to expect a negative correlation, because a large number of people in a small area could lead to a small value for blue areas per capita. In this case, the correlation coefficient is −0.32. This is (as expected) negative, but no linear relation was found; the size of blue areas per capita does not correlate with population density.

At the end of the statistical evaluation, the relations between both depicted parameters and the size of the town will be given:

• Main depicted parameter – proportion of blue areas (% of the total town area). The correlation coefficient is 0.088. It is thus evident that the proportion of blue areas is not dependent (as a linear function) on the size of the town.

• Supplementary depicted parameter – size of blue areas per capita. The correlation coefficient is −0.095, indicating that no linear relation exists between the two values. The size of a town thus does not correlate with the size of blue areas per capita.

If statistical dependency had been found in these two cases, it would have been difficult to arrive at an interpretation, and it would have been necessary to explore the causes of any such relationships. However, if neither of the depicted parameters correlates with town size, then town size does not need to be taken into consideration; this confirms that it was appropriate to graphically depict the cities as being of the same size.

The results of the analysis can be used in several different areas:

1. As a basis for detailed comparative analyses of individual towns. The results enable a particular town to be compared directly with another town (or a set of towns) with a similar proportion of blue areas or with an entirely different proportion, after which a detailed comparison can then be made with reference to other parameters.

2. For municipal authorities, as a basis for assessment and evaluation and as a starting-point for planning changes in the proportion of blue areas. Currently no similar comparison exists which would enable such assessments to be carried out. Given that blue areas are considered to be positive features in urban environments (from the perspective of the microclimate influence, aesthetic, recreational and other perspectives), a town’s ranking towards the bottom of the list could inspire the municipal authorities to take measures to increase the area of the blue areas in the town.

3. As a basis for town’ promotional activities, e.g. in tourism, when water surfaces are perceived as the attractive elements. Here a very interesting result is the first-place ranking of Cheb – a town which the Czech population does not commonly associate with water or blue areas in any way. The same applies to Karviná (third place) and Bohumín (fourth place), as well as other cities on the top of the list.

The use of one specific parameter does not enable generalizations to be made in other areas. For example, it is not possible to assume that a high proportion of blue areas means a high risk of floods, because flooding depends on topographic configurations which the study does not take into consideration.

6. Data visualisation

The aim was to visualise the results in a form of a Main Map, thus the format A0 was selected. Visualisation comprised the main part, additional maps, text and imprint field.

6.1. Main part

The basic visualisation method encompassed single-colour plan views of towns containing the plotting of blue areas. To allow the comparison, the towns were displayed in such a way that their areas were identical (for details see the Chapters 4.2 and 4.3). In the main field of the poster the plan views are distributed in a regular grid of the size of 10 × 10 fields. This part can be understood as 100 independent maps but due to the processing method used for the purpose of expressing quantitative characteristics, it is possible to range the method to the methods of thematic cartography. The classification of these methods is nonuniform (Miklín & Dušek, 2018), but generally the first step is a change in the area of the territory preserving its shape, which corresponds to a non-contiguous value-by-area cartogram (Miklín & Dušek, 2018; Tyner, 2010). The cartogram ensures that the location of the territory is preserved, which in the case of neighbouring states allows better orientation in the map. No advantage is associated with the preservation of the location in the case of separated selected towns, therefore the location was used to express another variable, which is the representation of blue areas; the towns were localised and sorted by this parameter. Because of impractical arrangement in a line, we opted for a grid or more precisely a system of ten lines. In the case of the methods not preserving the geographic location but whose ‘goals are clarity, simplicity, and speed of interpretation’ Tyner (2010) uses the term ‘diagram’.

When a town consisting of a few parts (mostly the central part and one or more exclaves) is visualised, the mutual geographic location of these parts is not preserved. Individual parts are visualised separately, yet the exclaves are pushed to the central part so that the town could be located in the regular grid without any huge gaps. Shifting has a positive effect on the perception of blue areas in the town as the town area makes an impression of greater compactness and clarity. The representation of blue areas is left absolutely unaffected by this method.

The colour of a town expresses the total area of blue areas (in m²) per capita in the town. Based on the recommendations of Brewer (1994) and Krygier and Wood (2005), we selected a graded colour scale from pale yellow to dark orange, which suitably contrasts with the blue areas. The scale consists of five follow-up intervals with borders based on quantiles.

Blue areas in towns are visualised using dark blue colour. Water bodies are displayed in a corresponding scale, water courses are displayed using a line of 0.35 mm or, if applicable, using the real width if it is greater that the given value. Narrow water courses are thus optically overestimated but this procedure is common in maps. The plan views of towns make it possible to identify water elements (bodies of water vs. water courses), river network density, and the distribution of blue areas in the town. It is evident whether the town was established on one bank of the river (Mělník, Uherské Hradiště etc.) or both the banks. The size of blue areas may be estimated as well, but considering irregular shapes of water bodies and overestimation of water courses, the estimate can hardly be precise. Therefore, this value was visualised under the name of the town – one blue point represents 0.5%. After the blue point marks there is given the precise value of the proportion of the blue areas in %. The size of a blue area per capita in m² can be identified from the colour of a town symbol.

The main part of the poster thus enables easy comparison of a hundred of most important towns from different points of view related to blue areas.

6.2. Additional information

The resulting cartographic work is complemented with a overview map of the Czech Republic showing the location of studied towns, main water courses and water bodies placed against a shaded relief. The map is supplemented with a simple scale and a north arrow, as recommended by Peterson (2009). Also, the localisation of the Czech Republic within Europe is included.

A common horizontal legend was added for the main and supplementary map according to principles presented by Slocum, McMaster, Kessler, and Howard (2005). Apart from the boundaries, the frequencies of towns are indicated for the intervals. At the same time, the legend serves for optical separation of plan views of towns and the accompanying text.

The Czech names feature a large number of diacritic marks above the letters (e.g. č, ř, á, ú, ů), and non-Czech (international) fonts do not always depict these diacritics with optimum clarity. For this reason, an original Czech font was chosen for town names and the other lettering in the maps. The font is Lido STF, designed by Czech typographer František Štorm. This font is based on Times New Roman, but it has been redrawn with the Czech diacritics redesigned to ensure maximum legibility while maintaining the quality required of modern graphic design (Štorm, 2000). The font has successfully been used in cartography to visualise the diversity of Czech Republic (Dušek & Popelková, 2017).

The text field at the lower edge clarifies what and how is represented.

The visualisation was designed to be legible on two levels (Krygier & Wood, 2005). From a longer distance, the headline is legible along with a general overview of all towns. Studying the map from a shorter distance, reveals details of individual towns, their names, presented values as well as the accompanying text.

Software

The data were displayed, processed and analysed using ArcGIS 10 software. MS Excel was used for the calculation of the studied characteristics given in the text.

Disclosure statement

No potential conflict of interest was reported by the authors.