Vol. 79 No. 2 Autumn 2006

GEOGRAPHIA POLONICA

POLISH ACADEMY OF SCIENCES INSTITUTE OF GEOGRAPHY AND SPATIAL ORGANIZATION Vol. 79 No. 2 Autumn 2006

GEOGRAPHIA POLONICA

POLISH ACADEMY OF SCIENCES INSTITUTE OF GEOGRAPHY AND SPATIAL ORGANIZATION WARSZAWA PUBLISHED by the Institute of Geography and Spatial Organization, Polish Academy of Sciences, Warszawa, Poland.

EDITORIAL OFFICE: Twarda 51/55, 00-818 Warszawa, Poland Tel. (48 22) 6978-858, Fax (48 22) 620-62-21. CONTRIBUTIONS and CORRESPONDENCE related to editorial matters should be sent to the Editor: [email protected] COPYRIGHT by the Institute of Geography and Spatial Organization, Polish Academy of Sciences, Warszawa, Poland.

SUBSCRIPTION INFORMATION: Geographia Polonica (ISSN 0016-7282) is published twice a year in Spring, and Autumn by the Institute of Geography and Spatial Organization, Polish Academy of Sciences.

SUBSCRIPTION ORDERS for Geographia Polonica can be placed with: ARS POLONA, Krakowskie Przedmieście 7; 00-068 Warszawa, Poland (including single past and current issues)

ISSN 0016-7282

INFORMATION ON THE CONTENT (from 1964) and forthcoming issues are available at: www.igipz.pan.pl/wydaw/home.htm

GEOGRAPHIA POLONICA IS CURRENTLY ON LINE! from vol. 75 The two latest issues (full text) are available at www.igipz.pan.pl/wydaw/GPonline.htm

PREPARED FOR PRINT by Aleksandra Deręgowska Department of Cartography and Geographic Information Systems, Institute of Geography and Spatial Organization, Polish Academy of Sciences, Twarda 51/55, 00-818 Warszawa, Poland CONTENTS

ARTICLES

MICHAŁ DOLATA and JACEK KOTUS—Social Production of Urban Space (A Case Study of ‘Bad’ Areas in Poznań) ...... 5

MICHAEL GENTILE—From Migration to Segregation in the Former Closed City ...... 23

MAGDALENA KUCHCIK—Defining Heat Waves—Different Approaches...... 47

ADAM ŁAJCZAK, JOANNA PLIT, ROMAN SOJA, LESZEK STARKEL and JUSTYNA WAROWNA—Changes of the Vistula River Channel and Floodplain in the Last 200 Years...... 65

BOOK REVIEW

Hungarian Spaces and Places: Patterns of Transition, edited by G. Barta, É.G. Fekete, I. Kukorelli-Szörényiné, J. Timár, Centre for Regional Studies Hungarian Academy of Sciences, Pécs, Hungrary, 2005, 595 pp.—DARIUSZ ŚWIĄTEK...... 89

REFEREES AND ADVISERS TO GEOGRAPHIA POLONICA 2005–2006...... 93

SOCIAL PRODUCTION OF URBAN SPACE (A CASE STUDY OF ‘BAD’ AREAS IN POZNAŃ)1

MICHAŁ DOLATA AND JACEK KOTUS Institute of Socio-Economic Geography and Spatial Management, Adam Mickiewicz University Dzięgielowa 27, 61-680 Poznań, Poland E-mails: [email protected]; [email protected]

Abstract: This article reflects upon the role of residents and social groups in assigning meaning to urban space. The research on which it is based was intended to: a) support a thesis about the importance of the social use of urban space in the shaping of urban structures, b) augment our knowledge as to the assignment of meaning to city areas and support the paradigm accounting for urban phenomena in terms of the everyday life of city dwellers, and c) identify the processes of urban marking in Polish metropolitan conditions by reference to criminal behaviour and its per- ception in Poznań. The research reported allows a more general conclusion to be drawn in that the stigmatization of urban areas only affects small areas and is fairly rare, while stereotypization of varying intensity is characteristic of much wider areas and is a more frequent mechanism un- derpinning social perceptions of places in a city.

Key words: Social production (marking) of urban space, stigmatization, stereotypization, Poznań city.

INTRODUCTION example, socially branded areas were, as if by laws of nature, endorsed by town coun- From the very dawn of the existence of towns, cillors, and located marginally in relation to it has been possible to identify a mosaic of the town centre understood in both its spa- areas in them inseparably associated with tial and social aspects (Wiesiołowski 1997). specified social groups. Artisans, clergymen, The executioner, ladies of easy virtue, the authorities, socially-isolated groups—they usher, and the lamp keeper in late-medieval all had a place in the spatial structure of the Poznań used a common area clearly marked medieval town. Because of its smaller scale, and separated both socially and spatially. its socio-spatial structures were easy to grasp, The relatively uncomplicated social struc- both in terms of their distribution within the ture of the then urban community involving town and their social characteristics. For only a few constituent elements of its spatial mosaic made it possible for those divisions 1 This article is an elaboration of the paper deliv- ered at the Seminar on Urban Knowledge (Łódź, April to be maintained in accordance with social 2004) and published as Kotus, J. and Dolata, M. (2004), norms, and with their administrative emana- Społeczne naznaczanie obszarów miasta [Social Mark- tion in the form of intra-urban location rules ing of Urban Areas] in Jażdżewska, I. (ed.), Konwersato- rium Wiedzy o Mieście, Uniwersytet Łódzki, 245–256. (Ihnatowicz et al. 1979). 6 Michał Dolata and Jacek Kotus

The present situation in small towns of centres there are office buildings, schools, Poland may be somewhat similar to that of and all kinds of housing arrangements. ages ago. A town’s layout and social charac- The contemporary city embraces many teristics are clear and simple. However, in areas that are socially produced (socially larger towns, spatial and social order at the marked) and hence perceived by residents, general urban level is practically a utopia, visitors and investors as good or bad places with smaller or larger social groups taking to live, enjoy themselves, or invest in. When over the use of various areas and giving them studying the results of such production, one various meanings. In large cities, the scale is should note whether the residents’ images of such that general urban phenomena fail to city fragments are being fixed by decisions be fully controlled by either city administra- of urban planners, and, as a result, whether tors or the urban community. Moreover, the they initiate processes of social and spatial socio-spatial structures of the largest Polish polarization (Hamnett 2001). cities are affected significantly by outside us- ers, as well as potential and actual investors, who contribute to the fragmentation of the THE ROLE OF RESIDENTS IN GIVING URBAN urban space (Martinotti 1996). SPACES SOCIAL MEANING Poland’s large cities are clusters of vari- ous, often disparate, areas located side by There is a clash of opinions among research- side. Within its administrative limits, each ers analyzing urban phenomena as to the city constitutes a world of its own in which it relations holding between the development is possible to identify spatially diversified so- of urban space and social behaviour. Some cial communities—communities using those emphasize the importance of the material- areas and through use giving them a specified -spatial sphere of the city, claiming it to be shape. One might say that, through those so- superior to the social sphere. The former is cial users, city areas become fully urban. This comparatively constant and rather difficult is to say, that the strictly spatial layer has feel- to change, in contrast to the city’s social ings engendered by being in different areas traits, which can be altered faster and more superimposed upon it, as well as opinions and readily. Hence social behaviour can be re- imaginings concerning them, assigned thereto garded as a result of spatial development by certain social groups. As Amin and Thrift (Wódz 1989). In an interesting study of the (2002: 8) put it, a city of socially produced formation of spatial barriers in a city intend- space is ‘an amalgam of often disjointed proc- ed to restrict freedom of use of various ar- esses and social heterogeneity, a place of eas, Flusty (1997: 48) lists five categories of near and far connections, a concatenation of space, which he characterizes briefly as: rhythms; always edging in new directions.’ • ‘Stealthy space—space that cannot be Large cities generate a heterogeneous found, is camouflaged or obscured by such community engaged in a multitude of social view impediments as intervening objects or activities, often very intensive in places. Of grade changes, key importance in this process are ‘street • Slippery space—space that cannot be communities’ (Jacobs 1961). It is those com- reached, due to contorted, protracted or munities of one street, one alley, one gate, missing paths of approach, or one corner that determine opinions about • Crusty space—space that cannot be a place by marking it with their behaviour. accessed, due to obstructions such as walls, Sometimes the opinions are restricted to gates and checkpoints, the place alone, sometimes they extend to • Prickly space—space that cannot be adjacent areas or an entire quarter. There comfortably occupied, defended by such are areas used by residents with a ‘good’ so- details as wall-mounted sprinkler heads ac- cial profile, and those controlled by social- tivated to clear loiterers or ledges sloped to ly marginal groups. Close to entertainment inhibit sitting, Social Production of Urban Space… 7

• Jittery space—space that cannot be their giving individual places a full shape, utilized unobserved due to active monitor- sometimes modifying their functions. Thus, ing by roving patrols and/or remote technol- social traits of a place can directly influence ogies feeding to security stations.’ a wider body of opinion about it, making it Naturally, the individual categories of flourish or perish. One can say that a socio- spatial limitations interpenetrate and occur logical answer to the question of what a city in various combinations, forcing the city- is and what mechanisms control it provides user to adapt his/her behaviour to the re- a basis for the consideration of many other quirements of a space. In various situations, urban issues (Mumford 1937/2002). the material endowment of an urban space is Lefebvre (1991, 1996) calls the social supposed to subdue spontaneous behaviour marking of urban locations the production on the part of its residents. The materiali- and reproduction of places in the city. ‘Ang- stic paradigm imposes the view of the city as lo-Saxon’ scholars speak of (re)negotiation of a holistic system superior to its inhabitants the meaning of areas and their (re)structuring (Geddes 1915). under the influence of user groups (Mooney However, other scholars feel there is no 1999). With the help of these terms, in a the- doubt about the social roots to the genera- oretically ideal situation, the process of city tion of space or endowment of some areas structuring and social participation can be with certain social meanings (Wallis 1979; said to go through the following stages: Jałowiecki 1988; Lefebvre 1996; Allen 1999; • The stage of planning the new city (or and Pile 1999). Llewellyn (2004: 230) claims, new objects therein, e.g. housing estates, en- after Lefebvre (1996), that ‘space is not tertainment parks, green space), merely produced for simple consumption, • The stage of introducing new users but that spaces can be adapted, manipulat- into the city (into the new objects), ed, appropriated and produced by a range • The stage of assigning meaning to (re- of individuals.’ This can be called a dispute negotiating meaning in) the city (to the new between architectural and social determin- objects), and ism, between subjective and objective mean- • The stage of restructuring of the city ing of space (Herbert and Thomas 1990), or its areas, as a result of the social renego- between traditional urbanism and new ur- tiation of meaning. banism (Amin and Thrift 2002). ‘Neotra- In search of examples of this process ditional urbanism seeks to provide qual- among cities built from the ground up with ity public spaces that are semi-enclosed and a certain social mission that has been altered legible, and that connect places that people in social practice, one may mention the well- use, in contrast to the amorphous, illegible, known case of Brasilia (Holsten 1989; Bau- isolated and largely unused public spaces of man 2000). Brazil’s capital, designed to be the master-planned community’(Ellin 1997: a social urban ‘Eden’, with ideally planned 30). Socially produced and reproduced are neighbourhood areas and places for various both ‘indoor residential settings’ and ‘out- group activities, has become an example of door residential spaces close to home,’, while the unpredictability of social movements such territorial markers as ‘swept sidewalks, and the social impotence of town planners. scrubbed steps, trimmed bushes, ... pink fla- A Polish example of such a town is Borne mingos, pottery cats […] convey information Sulinowo. During the Second World War it to neighbors, passers-by and would-be bur- was a closed unit housing German soldiers. glars’ (Delaney 2005: 50). After the war the town’s status did not There is no doubt that cities offer their change, except that it was now occupied by residents material-spatial structures that are the Soviet Army (the last soldiers of what was indeed more permanent than social phe- by then the Russian Federation left in 1992). nomena. However, the very fact of people For almost half a century, then (untill Janu- moving into a developed urban area involves ary 2003), the town was practically closed 8 Michał Dolata and Jacek Kotus to civilians. Thereafter its character started stigmatization and the authorities’ silent ap- to change, and the residents and authorities proval. Thus, in the 1970s the Soho area be- began adjusting the military infrastructure came dominated by the pornographic indus- to the needs of a civilian population. While try and earned the reputation, also abroad, the origin of this particular settlement unit as the centre of London’s sex industry and is different, one might assume that the ‘new’ a ‘red-light district.’ history of Borne Sulinowo started when the It was only wide-ranging measures insti- Soviet troops had left it, and observe how the tuted by the local and central authorities that new users have been changing the functions arrested the process of the negative stigma- of individual places in the town in a way tization and stereotypization of Soho (Hub- common to the entire settlement unit. bard 2004). The creation and use of ‘red-light In most cases, however, the process of the districts’ in British inner cities is an explicit social renegotiation of the meanings of plac- example of the social stigmatization of a city es and their restructuring is limited to cer- by its users (Hubbard and Sanders 2003). tain areas, e.g. a park or fragment of a hous- Sometimes, however, city authorities ing estate (Llewellyn 2004). A spectacular have initiated measures reinforcing the example of appropriating places and giving stigmatization of places on their territory. them meanings is described in Young (2003: That is what happened in Chicago with the 617). Studying the spatial behaviour of street launch of the North Halsted Streetscape children in Kampala (Uganda), he noted Project, under which areas occupied by eth- a tendency for niches to be created in mar- nic minorities received small-architectural ginal spaces. A variety of places occupied accents: miniature copies of temples, monu- created ‘untouchable spaces’, ‘underground ments, gates to the quarters, etc. (the Gre- spaces’, or ‘rooftop spaces.’ Those places ektown Project, Chinatown Project, Puerto came to function as homes and playground, Rico Project). In this case, the administra- and these new functions were accepted and tive measures were a consequence of earlier authorized by other inhabitants. At times, social behaviour and were intended to meet however, persons with different ideals or social needs (Reed 2003). different place-images of city life made an A short but unequivocal description of effort to prevent the children from appro- such a process in Poland is given by a Pol- priating public and neighbourhood areas, ish sociologist Wódz (1989) in respect of the and to prevent those areas from acquiring so-called ‘Nowa Huta Scarp’. Nowa Huta a bad name. is a district of Cracow built in the 1950s as Sometimes the appropriation and social a huge housing estate for workers of a me- marking of urban places by certain social tallurgical plant. Among the many blocks of groups combine with conflicts over neigh- flats there was an undeveloped piece of land bour and public spaces to force local au- on a scarp. Groups of residents spent their thorities into legal and/or administrative ac- free time there marking it socially. The au- tion. When prostitutes in Portland’s central thor distinguishes three stages in the social streets became a nuisance to other everyday production of the scarp space: ‘no-man’s- users of the same areas, the authorities felt -land’; start of ‘occupation’ by residents; and it their duty to issue a Prostitution-Free ‘established’ patterns of behaviour. Zone Ordinance, in which they delimited a ‘zone of exclusion’ and thereby removed the women from the city centre. In this way BETWEEN STIGMATIZATION AND the local authorities counteracted its social STEREOTYPIZATION OF URBAN AREAS marking (Sanchez 2004). Hubbard (2004) describes changes in the area of London’s The social renegotiation of meaning as re- Soho, which had become an informal ‘pros- gards places in a city can lead in some cases titution tolerance zone’ as a result of social to their social stigmatization or stereotypiza- Social Production of Urban Space… 9 tion. The term stigmatization can be found live in them. A fixed opinion may sometimes in the professional literature, although it have nothing to do with the facts (Aronson is not defined explicitly (Mooney 1999: 81; et al. 1994). Wacquant 2002: 225). The present article Wódz (1989) notes the very bad opin- thus seeks to endow it with a more precise ion Nowa Huta acquired for itself among definition, and to elucidate the relationship Cracovians in the 1960s and 70s. The between it and stereotypization. opinion embraced the entire area, which The urban dimension to stigmatiza- was talked about in terms of it being a ‘bad tion is initiated by the appropriation and space’ or ‘bad neighbourhood’. The Craco- use of certain places in a town by specified vians did not distinguish any particular social groups. This concerns both individ- ‘bad’ places within the district. In reality, ual objects like a pub and larger areas like however, Wódz’s research showed that ‘de- a good or bad neighbourhood (Mellor 1985; viant behaviour’ was characteristic of only Campbell 1993; Robins and Aksoy 1995, one area in it. He suggests that the wide- 1996). What is important, on this reading of spread opinion as to the neighbourhood social stigmatization, is that the opinions of being ‘bad’ stemmed from the presence residents about some places in their city are within it of particular places that were true, resulting from a knowledge of places egalitarian and devoid of any spatial dis- plus groups ascribed to them and constitut- tinctions. This is certainly an interesting ing a confirmation of the existing state of observation as regards socially stigmatized affairs. areas, in the sense that it portrays stig- The social stigmatization of urban places matization turning into stereotypization. can sometimes be the first stage of a broader If respondents call entire districts bad process of their stereotypization, or the as- (Jałowiecki 1980; Biderman 1999), this signment to them of meanings under the could be deemed a clear example of stig- influence of simplifications and errone- matization-turning-into-stereotypization ous and or superficial social beliefs. Beliefs of urban areas. Finally, opinions about about places that become fixed and pass into places in a city may emerge as totally out public opinion are sometimes derived from a of touch with reality, or reflect an old, his- true source, but often also from a source that torically distant past. This situation is ter- has long ceased to exist (Figure 1). In such med full stereotypization. cases, city areas will be neither as good nor In summing up the above reflections as bad as they are commonly reported to be, about stigmatization and stereotypization and the same will be true of the groups that of certain urban areas, we can restate the

Figure 1. The mechanism of the transition of a different use of space into a socially stigmatized and stereotyped one. Source: Authors’ own compilation on the basis of Kotus, J. and Dolata, M. (2004). 10 Michał Dolata and Jacek Kotus three dimensions to the phenomenon, i.e. solely the product of beliefs based on old, (Figure 2): almost historical, facts, or when the opin- • Stigmatization: when the opinion about ion is entirely groundless. a place and the group using it coincides with Social production involves both actual the actual situation, users initiate social stig- (stigmatization) and stereoptypic (stere- matization of the place. otypization) attribution to urban areas of • Stigmatization turning under the specific social characteristics that become influence of stereotypization from full to fixed in the consciousness of city dwell- partial: when the opinion about a place ers. The observers ( Figures 1, 2) are the and its group of users is exaggerated in re- residents of the thus-produced areas, or lation to reality, the area that social opin- of other parts of the city who perpetuate ion extends to is larger than that in which their image in public opinion and collec- the phenomenon actually occurs. tive awareness. • Full stereotypization: when the opin- ion about a place and its group of users is

Figure 2. Dimensions to the social marking of urban areas —The perspective of an observer. Source: Authors’ own compilation on the basis of of Kotus, J. and Dolata, M. (2004). Social Production of Urban Space… 11

‘DANGEROUS PLACES’: THE CASE est-esteemed quarters on the residential OF THE SOCIAL MARKING OF CERTAIN map of Poznań. Jeżyce and Łazarz were AREAS IN POZNAŃ2 then much sought-after as housing locations and residence there was highly appreciated As negative phenomena are often a source of (Skuratowicz 1999). Another category of very strong opinions, they are a good back- areas seen by Poznań residents as notorious ground to descriptions of place marking in includes the large block estates of Piątkowo, cities. In addition, the identification of really Winogrady and Rataje; and a third includes bad places (those stigmatized through their the extensive green spaces within the city users) and their differentiation from those limits (Cichocki and Podemski 1998, 1999). that are bad by virtue of a stereotype allows Each of these extensive areas is judged in for the instituting of various measures to im- general terms, without resort to the indica- prove the image of the urban space and cre- tion of specific places. The bad opinion can ate a community enjoying a sense of safety thus be assumed to be largely stereotypic, (Pain 2000, Pain and Townshend 2001). The especially in the case of Jeżyce and Wilda, presentation of the processes of stigmatiza- which are actually two extensive, and highly tion and stereotypization of places in the diversified districts of Poznań. The all-inclu- city of Poznań will proceed through three sive bad opinion and perpetuation of a ste- stages: reotype is also revealed in the assessments • Presentation of the negative popu- of the block estates and green spaces. The lar opinion about certain places in Poznań reasons for the areas in question being con- (stereotypes) and reasons why the Poznań sidered bad by respondents are (Biderman residents can regard those places as bad 1999: 1001, Cichocki, Podemski 1998: 91): (stigmas), • that the areas with old tenement hous- • Presentation of the results of a research es support a high level of social deviance and on crimes and offences in Poznań, and no safety, while the buildings are old and di- • Comparison of those results with the lapidated and the inhabitants ‘bad’, opinions of city residents—as research par- • that the block estates sustain a higher ticipants—in regard to their feelings of safe- level of anonymity and no safety, ty, and their perceptions of certain places as • that the green spaces are gathering dangerous. places for hooligans offering no safety. Naturally, there is no defending the the- STEREOTYPES AND STIGMAS CONCERNING sis that all inhabitants of old tenement hous- PLACES IN POZNAŃ THAT ARE BAD es manifest social ills and are marginal, that IN POPULAR OPINION all quarters are dangerous, that each major In the popular opinion of Poznań residents, green area is threatened with delinquency, Wilda and Jeżyce, together with their subsec- and that residents of block housing estates tions: Łazarz and Stare Miasto (Old Town) 3, live in a permanent sense of danger. The have long been regarded as bad districts opinions revealed by the respondents are (Cichocki and Podemski 1998; Biderman clear generalization of a stereotypic nature. 1999). This means that there is a category In looking for the roots of such views, of city quarters of ill repute which includes the most one can find are ‘bad’ city blocks, those with a predominant building pat- fragments of streets, or possibly even single tern of old tenement houses or with a large building gates, in which socially deviant be- number of such areas. Interestingly enough, haviour can be observed. It is thus possible a century or more ago these were the high- to advance a thesis, which will be elaborated in more detail in subsequent sections, that 2 One of Poland’s largest cities with close to 600,000 such negative stereotyped opinions are usu- inhabitants. It is a regional centre of industry, culture, education and science. ally underpinned by a grain of truth as re- 3 For their location see Figs. gards a concrete place at which criminal acts 12 Michał Dolata and Jacek Kotus tended (or still tend), to occur more often measured by the number of incidents re- than usual, or where such acts are merely ported per 10,000 head of population. This suspected of the group using the place. For index (at 179.2) was three times higher than example, the old parts of the city may indeed the city average, though should be treated feature groups of younger or older people with caution, since the numbers present in loitering around gates, sometimes drinking the city centre every day are higher than alcohol or teasing passers-by; the same holds that of permanent residents. The phenom- true for the green spaces. Certain block es- enon of population concentration in the tates do in turn include places appropriated downtown part of a city during the day, as by young people, who like to meet there and swelled by the inflow of people from the sometimes also drink alcohol or harass oth- periphery of the metropolitan area is well er users. There have been individual cases known. In terms of the kinds of offence, of obvious vandalism or crime. We are thus as broken down by day and night, the Old faced with typical examples of places being Town did not depart from the city average stigmatized by reference to the social groups with its 61% of crimes taking place in day- using them. The negative traits of the users time. The precinct was peculiar in terms of are noted, fixed in public opinion via labels, the kinds of act perpetrated there, however. imposed on such places and, with time, ex- Unlike other sectors of the city, it had an un- tended to cover a larger area.There is no usually high rate (53%) of acts committed in doubt that this negative stigmatization turn- typical public space, i.e. thefts in shops and ing into stereotypization is greatly enhanced pubs; street robberies, thefts and battery; by the media, which publicize individual night disturbances, and misdemeanours of ‘bad’ cases and make generalizations. groups of young people. On the other hand, it had the lowest percentage (25%) of offen- AREAS OBJECTIVELY THREATENED ces involving cars. BY CRIME In the residential quarter of Wilda, which The identification of areas within Poznań is old, densely built up and of ill repute, the variously threatened by crime was based on rate of crime (offences per 10,000 head of an analysis of crimes and offences recorded population) over the study period amount- by the police in March 2003.4 In objective ed to 75.2, which was markedly higher than terms (the number of incidents), the threat of in the neighbouring precinct of Grunwald crime is markedly higher during the day than (49.6). Wilda had a very high percentage of at night (Figure 3). The particular categories daytime crime (70%), while in terms of the of offence extend across the whole of Poznań, kind of offence, those involving public space despite differences in rates (Figure 4). predominated (38%). Next came those in- There are some police precincts, in which volving cars (32%), most of which (unlike in certain kinds of offence are more frequent. the Old Town), were thefts of or damage to For example, the Old Town precinct reports cars near the houses of their owners. the highest rate for crimes and offences as As has already been observed, the po- lice statistics make it clear that a Poznań 4 The presented results derive from an analysis, citizen can fall victim to crime in most made in 2003, of materials found among the records parts of the city, with particular places dif- of incidents and police interventions at particular po- lice stations (PS) of the Municipal Police Department fering only in the rate and locally predomi- (MPD) in Poznań. The available statistics on crimes and nant kind of crime. However, each precinct offences were chiefly presented by station, whose pre- has trouble spots in which a larger number cincts coincide with the limits of the city’s administrative districts. An exception is the Old Town district, which of offences are concentrated. These are has two police stations: Poznań-Old Town and Poznań- not spatially indeterminate and often have North. The month in which criminal acts were commit- easily delineated ranges. For example, the ted was chosen at random, because it was impossible to obtain material for the entire year. The statistics only common belief that block estates generate cover events reported. the kind of deviant behaviour associated Social Production of Urban Space… 13

Figure 3. Offences in Poznań during the day and night, by police station, March 2003. Source: Authors’ own compilation on the basis of police data.

with hooligan groups, is not borne out by From the above three conclusions are the fact that (with three exceptions) most relevant to reflections on social marking in Poznań estates of this type do not display a city: an increased incidence of offences. The • most acts are committed by day, exceptions, for which the data collected • precincts differ as regards the rate reveal a prominent place among police and kinds of crime, but all areas are at records are the black areas in Figure 5. risk, The situation is similar in Jeżyce precinct, • there are no especially crime-ridden in which the majority of criminal acts con- districts, only places (streets, squares, es- centrate in the few streets forming ‘Old tates) in each pricinct with increased inci- Jeżyce’. The aforementioned Wilda, seem- dences of offences. ingly dangerous throughout, also in fact These conclusions were set against opin- has individual streets and squares that are ions of Poznań residents regarding their ex- especially unsafe and thus give the whole posure to crime. district a bad name. Places in Wilda most exposed to crime include, as in the Stare AREAS PERCEIVED BY POZNAŃ RESIDENTS Miasto (Old Town), both busy streets with TO BE EXPOSED TO CRIME shops and services (Wierzbięcice, Górna A survey of public opinion on levels of crime Wilda) and less busy ones on the periphery in Poznań and on the spatial distribution (Figure 5). thereof, was carried out in 2002. It involved 14 Michał Dolata and Jacek Kotus

Figure 4. Offences in Poznań during the day and night, by kind and police station, March 2003. Source: Authors’ own compilation on the basis of police data.

457 city residents5 representing three sec- ing pattern is of multi-family blocks of flats tions distant from one another and differing (34.2% of respondents), and the Bajkowe in town-planning and architectural terms. (Fairytale) Estate, a suburban quarter with These were: the old part of Jeżyce within the predominantly single-family houses (33.0% city centre, as featuring a compact building of respondents). pattern of old tenement houses (32.8% of The respondents were asked to assess respondents), the Pod Lipami (Lime Tree) safety in Poznań, to indicate especially dan- Estate in the city centre, in which the build- gerous places in their immediate neighbour- hood and in more distant parts of the city, 5 The sample was of 58.7% female and 41.3% male; and to appraise their reliance on sources of 4.4% persons had elementary education, 16.4% voca- opinion about dangerous places. It should be tional, 45.9% secondary, and 33.3% higher; there were 27.1% persons aged 18–25, 24.3% aged 26–39, 16.8% emphasized that most respondents had ne- aged 40–49, 16.4% aged 50–59, and 15.4% were 60 and ver been victims of crime themselves. Only over. 10% had suffered as a result of an offence Social Production of Urban Space… 15

Figure 5. Places in Poznań with the highest rates of selected categories of offence, March 2003. Source: Authors’ own compilation on the basis of police data. committed in their flats, houses or their vi- neighbourhood and in public places (3.9 and cinity; 14% in the nearest neighbourhood of 3.7, respectively). They feel most threatened their places of residence; and about 20% in after dusk, and more so in public places a more distant public space. Often enough, (2.4) than near home (below 2.8)(Figure 6). the same people were involved. This offers In view of the facts presented these results sufficient grounds for the conclusion that the justify the statement that the perception of respondents formed their opinions largely the threat of crime in terms of day and night on the basis of all kinds of reports (prejudic- is false. It reveals a stereotyped valuation es and stereotypes prevalent in the Poznań of city sections following on from the con- urban community), rather than their own viction that it is more dangerous after dusk experience. and at night, and in public places and parks. The assessment of the feeling of safety In reality, the rate of daytime crime is higher, showed it to be clearly dependent on two irrespective of whether we are dealing with factors: the time of day and the distance a neighbourhood or spaces of community life. from the place of residence6, with the former The following points can be made when seeming to be more important. The respon- comparing the above opinions of the re- dents feel safest in their homes (a mean of spondents in terms of their types of residen- 4.3) and in the daytime in both their own tial quarters: • There are no major differences in the 6 The assessment of the feeling of safety in the city assessment of the safety level near home by ranged from 0 (highly unsafe) to 5 (very safe). The re- day. Differences do, however, appear, re- spondents had to consider six categories: (a) staying at home, (b) walking near home in daylight, (c) being in garding walks after dusk, which were mainly public places in the daytime, (d) being in parks, (e) walk- judged dangerous by inhabitants of old ten- ing near home after sunset, and (f) being in public places ement houses (over 60% of responses as after sunset. against 47% of residents of the block estates, 16 Michał Dolata and Jacek Kotus

Figure 6. Evaluation of the sense of security in the city (arthmetic mean: 0 – very unsafe, 5 – very safe) Source: Authors’ own compilation on the basis of poll data.

and 38% of those living in single-family The survey was absolutely positive about houses). Parks are another category which the actual occurrence of dangerous places elicited differences of opinion: they scored in Poznań. As many as 75% of respondents the lowest with residents of single-family stated that there were such places in the city, houses, a group of respondents that also felt while a mere 2.5% claimed there were none. less safe in public places by day and night. The dangerous places were to be found pri- • Public places after sunset tended to be marily in ‘other parts’ of the city, i.e., not seen as dangerous by older people, as well in the respondents’ places of residence or as by the less-educated, while the youngest their immediate vicinity (57%). Only 15% age group (18–25) perceived them as fairly claimed dangerous places were located safe. People with higher education were am- mainly in their own neighbourhood. This bivalent. Women were also more inclined clearly suggests that respondents were more than men to be afraid (over 80% as against prejudiced against outside areas, in other 45%). words those that were less known to them. • In the assessment of safety in public The assessment as to where dangerous plac- places by day, in contrast, gender was no dis- es are located is equally interesting when criminating factor. Those with higher educa- analyzed by the respondents’ places of resi- tion and younger tended to think better of dence, viz. old tenement houses in the city this than those with elementary and voca- centre (Jeżyce), multi-family blocks of flats tional education and of advanced age. (the Pod Lipami Estate), and single-family • Neither gender nor age were discrimi- houses (the Bajkowe Estate)(Figure 7). nating factors when it came to the assessment Residents of the single-family Bajkowe of safety near home in the daytime. The re- Estate considered that such places were spondents who felt safer in those areas after mostly located outside their own neighbour- dusk were younger, with higher education, hood, in other parts of the city. This opin- and male. ion also prevailed among block residents, Social Production of Urban Space… 17

Figure 7. Location of especially dangerous places in the opinion of residents of particular areas Source: Authors’ own compilation on the basis of poll data. though a higher proportion than in the pre- • were less likely to claim that there were vious group (13.5%) believed that they could fewer offences in their neighbourhood than be found mainly in the neighbourhood. It anywhere else. was the group of tenants of old houses that The people living in blocks of flats on was most pessimistic about their place of the Pod Lipami Estate largely declared that residence: fewer responses pointed to other there were fewer incidents of aggressive parts of the city as dangerous. behaviour, battery and theft in the estate Moreover, residents of the single-family streets than anywhere else. This is interest- Fairytale Estate were more likely than the ing in account of the prevalent opinion in other respondents to claim that: Poland, fuelled by the mass media, that block • there were rather more breakings on housing estates are often controlled by gangs their estate than elsewhere in the city, of juvenile delinquents, so-called ‘blockers’. • there were decidedly no cases of ag- As the poll showed, these negative stereo- gressive behaviour in the street, types were not corroborated by respondents • there were no, or decidedly fewer inci- living in one of the areas in question. dents of battery and theft in the street, The analysis of sources of opinion on • damage to, and thefts of, cars were dangerous places in Poznań confirms that more frequent there than anywhere else, respondents usually form their statements • there were none of the offences under on the basis of hearsay. In the case of 72% of study on their estate, or decidedly fewer than respondents, statements were based on opin- in other parts of the city. ions of friends and other people. A signifi- The tenants of the old houses (Jeżyce): cant role was also played by the press and te- • felt there were more cases of aggres- levision (65% of respondents relied on them, sive behaviour in the streetwhere they lived ‘definitely’ or ‘rather’, in forming their opi- than anywhere else, nions). A decidedly smaller opinion-forming • felt there were more incidents of bat- effect was ascribed to police reports and the tery and theft in the street, respondents’ own negative experience (44% • felt there were more thefts of cars, and 39%, respectively) (Table 1). 18 Michał Dolata and Jacek Kotus

Table 1. Reliance on sources of opinion about dangerous places in the city (missing data are omitted)

Categories of reliance Sources of opinion about dangerous places Definitely yes Rather yes Rather not Definitely not Net indicator Own negative experience 20.1 18.8 32.2 15.9 -9.2

Press and TV 18.8 46.4 18.2 5.0 42.0

Friends and other people 22.9 49.5 12.0 3.3 57.1

Police reports 14.0 29.9 27.7 15.9 0.3

Source: survey

The tenants of old houses usually claimed residents produced the following results that their opinions on places of danger relied (Figure 8). In their own neighbourhood, on their own negative experiences and opin- the Jeżyce inhabitants chose mainly Jeżyce ions of friends. Inhabitants of single-family Square and Wawrzyniaka Street (areas with houses mostly relied on media information. old tenement houses). The former is a typi- Women were more likely to rely on news cal public space and the district’s principal from the media (nearly 80%) than men (un- centre of small street trade. Densely packed der 60%). There were also about 10% more in the surrounding frontages are services, women who gained their knowledge in this chiefly shops. During the day it is one of respect from friends. people with lower edu- the busiest places in the city. Wawrzyniaka cation largely relied on their own negative Street, in turn, is a relatively short (550m experiences, while those with higher educa- long), rather quiet street with few character- tion tended to indicate the indirect sources istic features, lined by old tenement houses. of information, viz. the press, friends, or po- In the light of the statistics, the opinion of lice reports. the Jeżyce respondents concerning the thre- A qualitative analysis of places indicated at of crime in their district is only half ju- as dangerous by the three groups of Poznań stified. While Wawrzyniaka Street indeed

Figure 8. Dangerous places in Poznań in the opinion of residents of the Pod Lipami and Bajkowe housing estates and of the Jeżyce area. Source: Authors’ own compilation on the basis of poll data. Social Production of Urban Space… 19 has a high incidence of criminal acts, Jeżyce and reflect a correct appraisal of the central Square can hardly be counted among the part of thecity as an area with an above-av- especially dangerous areas. Its bad opinion erage incidence of crime. The opinions did may partly stem from the fact that the emp- not indicate precisely which places were dan- ty stalls are nocturnal meeting places for gerous, but conveyed negative views about groups given to alcohol consumption. a wider area. Among dangerous places in their neigh- bourhood, respondents from the Pod Li- pami Estate (of blocks of flats) primarily BETWEEN STIGMATIZATION mentioned the nearby green spaces, includ- AND STEREOTYPIZATION—CONCLUSIONS ing the immediately adjacent Gagarina Park and the Citadel across the street (mentioned Given the conclusions to the theoretical in two-thirds of all responses) (Figure 8). reflections presented earlier, the authors However, in none of those areas did the po- used the results of their research to identify lice record an increased incidence of crime cases of stigmatization (whether full or par- over the study period. On the other hand, it tial) and stereotypization in the area under is significant that only a small proportion of analysis. the respondents indicated other compara- An undoubted example of full stigma- tively close areas, like the B.Chrobry Estate tization is Wawrzyniaka Street in Jeżyce (also of blocks of flats), which is known to district. Worth noting is the fact that it lies feature an above-average incidence of crimi- in respondents’ neighbourhood. Hence the nal acts. very precise identification of an area which The Bajkowe Estate respondents gave the shows an increased incidence of criminal Marcelin Wood as an especially dangerous acts and functions as dangerous in social area in their neighbourhood (referred to by consciousness. 66%!) (Figure 8). This is a small (ca. 200 ha) The analysis also makes it possible to in- stretch of municipal woodland, hardly at- dicate cases of stereotypization in the city. tractive in landscape terms and poorly devel- Between stigmatization and stereotypiza- oped, used by the inhabitants of the nearby tion a certain continuum is assumed to ex- estates for walks. It is crisscrossed by the ist, in the form of partial stigmatization or shortest footpaths and cycle paths joining stigmatization changing into stereotypiza- the Bajkowe Estate and Ławica with the cen- tion (Figure 2). That there were stereotypes tral part of the Grunwald district. As in the of areas was revealed by respondents from case of the Gagarina Park and the Citadel, the Pod Lipami and Bajkowe Estates. They we are here dealing with respondents’ false expressed their opinions about the Citadel images of the scale of crime in their imme- and the Marcelin Wood as places especially diate neighbourhood. The Marcelin Wood exposed to criminal acts, but the opinions does not by any means figure as a ‘black are not corroborated by known facts as re- beat’ in the local police records, in terms of flected in police records. Detailed qualita- the crimes and offences associated with it. tive research might lead to the uncovering Among dangerous places in other (non- of the genesis and strength of those stereo- neighbouring) sections of the city indicated types. It is impossible to establish what has by the respondents from Jeżyce, and the made the park and the wood earn the opi- Pod Lipami and Bajkowe Estates, the most nion of high-risk places. Perhaps the senti- prominent was Wilda (26–37%), especially ment originated in some isolated event (like the old section of the district with the rail- a vicious murder or rape) which, although way and coach stations. Next came other dating to a rather distant past, has been fixed downtown areas: so-called Old Jeżyce, the in the collective consciousness of the local Old Town, and the oldest part of Grunwald, communities and passed on in the form of a Łazarz. These opinions are closer to reality bad opinion about the place. In both cases, 20 Michał Dolata and Jacek Kotus after all, we have areas with trees and shrubs, Piątkowo (the Poznań-North precinct). Both by their very nature isolated, less frequented, these areas are inhabited by a total of some convenient for a potential criminal as a place 250,000 people. They have come to be ge- of action and refuge. It is also highly prob- nerally regarded as an arena of a ‘new’ type able that those places are a summer haunt of crime manifesting itself primarily in of- for groups of alcohol drinkers, which, in the fences committed by youth groups represen- opinion of observers, may be associated with ting the sub-culture of so-called ‘blockers’. potential crime. And indeed, as follows from police records, There is also partial stereotypization in some large block estates (the B.Chrobry, the case of the district of Wilda. As has been Stare Żegrze, and the Orła Białego Estates) mentioned, this section of the city does in- stand out for their incidence of criminal acts. deed feature an above-average incidence of However, in most of them misdemeanours selected categories of crime. They include by juvenile delinquents rarely define broad- primarily daytime offences committed in er spaces. Rather, the groups tend to mark the so-called public space (robberies, thefts small sections of the estate space, and the and misdemeanours in the street, shops and occurrence of an offence in any particular eating places). The district was indicated spot is hardly predictable. Thus, what we are as dangerous by a substantial percentage of dealing with here is a situation in which the the respondents, and not only by those liv- area covered by social opinion (all the large ing relatively close to it (Jeżyce and the Pod block-of-flats estates) is greater than that in Lipami Estate), but also by residents of the which the phenomenon concerned actually more distant section of the city (the Bajkowe occurs, i.e. with partial stigmatization. Estate). In this case social opinion marks It is possible to draw a more general the entire district (‘dangerous Wilda’, ‘bad conclusion from the research reported: the Wilda’), while the groups making reference stigmatization of urban areas only affects to it display characteristics that may or may a small area and is fairly rare, while stere- not favour marginal phenomena. The users otypization of varying intensity is charac- that initiated the negative social stigmatiza- teristic of much wider areas and is a more tion of the area turning into stereotypization frequent mechanism of social perception of are the Wilda inhabitants. They derive part- places in the city. ly from socially and economically degraded Phenomena involved in the social pro- groups of the old proletariat working in the duction of urban places are not easy to local heavy industry. In recent years, how- control. They are spontaneous, result from ever, a new group of residents has emerged psycho-social processes of perception, are as a result of the appearance of new hous- difficult to anticipate, and develop over a pe- ing estates that are monitored and closed. riod of many years or even decades. Hence it They are inhabited either by people of better is practically impossible to avoid stigmatiza- material status and good jobs, or by students tion and stereotypization of places in a city. renting flats in a section of the city close to In itself, the social production of urban the centre. In the opinions of the respon- places analysed in the present article is not dents, the latter also contribute to the ‘bad a negative process. These mechanisms cause Wilda’ image, notwithstanding the fact that urban spaces to acquire social meanings and this is far from being the true situation. stop being physical locations. What is negati- Stereotypization also occurs on block ve is the transformation of the opinion abo- housing estates, which are elements typical ut certain places into a standard that guides of the spatial structure of all large cities in decision-makers, investors, planners, and Central and Eastern Europe. The Poznań outside users, e.g. tourists. Such mechanisms districts with such estates are the afore- can be dangerous, in that they may reinforce mentioned Rataje (the Nowe Miasto (New the processes of urban polarization, whether Town) district) and Winogrady together with at the level of individual places (like housing Social Production of Urban Space… 21 estates), categories of place (parks at night, Delaney, D. (2005), Territory. A Short Introduc- block estates), districts, or larger sections of tion, Blackwell Publishing, Oxford Malden. a city. The division into good and bad places Ellin, N. (1997), Shelter from the Storm or Forms determined by social opinion, whether on the Follow Fear and Vice Versa, basis of observation and reports of the actual in Ellin, N. (ed.) Architecture of Fear, Princeton behaviour of user groups or on the basis of Architectural Press, New York, pp. 13–46. stereotypes, can be of consequence to town Flusty, S. (1997), Building Paranoia, in Ellin, planning. The prevailing good or bad opi- N. (ed.) Architecture of Fear, Princeton Archi- nion may affect actions by the local authori- tectural Press, New York, pp. 47–60. ties and the investment climate. This, in turn, Geddes, P. (1915), Cities in Evolution, Williams may lead to the location of investment in os- and Norgate, London. tensibly good places, and hence to a further Hamnett, C. (2001), Social Segregation and Social decline of those commonly considered bad. Polarization, in Paddiso, R. (ed.) Handbook of Urban Studies, Sage, London,Thousand Oaks, New Delhi, pp.162–176. ACKNOWLEDGMENTS Hampton, K.N. and Wellman, B. (1999), Netville On-Line and Off-Line. Observing and Sur- The authors would like to thank the two veying Wired Suburb, American Behavioral anonymous reviewers for their valuable Scientist, 43, 3: 475–449. comments and ideas which have helped to Herbert, D.T. and Thomas, C. J. (1990), City in improve this article. Space, City As Place, David Fulton Publisher, London. Holsten, J. (1989), The Modernist City: an Anthro- REFERENCES pological Critique of Brasilia, Chicago Univer- sity Press, Chicago, Ill. Amin, A. and Thrift, N. (2002), Cities. Reimagin- Hubbard, P. (2004), Cleansing the Metropolis: ing the Urban, Polity, Cambridge. Sex Work and the Politics of Zero Tolerance, Allen, J.(1999), Worlds within Cities, in Massey, Urban Studies, 41, 9:1687–1702. D., Allen, J. and Oile, S.(eds.), City Worlds, Hubbard, P. and Sanders, T. (2003), Making Space London, New York, Routledge, 53–98. for Sex Work: Female Street Prostitution and Aronson, E., Wilson, T.D. and Akert, R.M. the Production of Urban Space, International (1994), Social Psychology, the Heart and Mind, Journal of Urban and Regional Research, 27,1: Harper Collins College, London. 75–89. Bauman, Z. (2000), Globalizacja [Globalization], Ihnatowicz, I., Mączek, A. and Zientara, B (1979) Państwowy Instytut Wydawniczy (PIW), Społeczeństwo polskie od X do XX wieku Warszawa. [Polish Society from the 10th to the 20th Cen- Biderman, E.(1999), Geografia społeczna i poli- turies], Książka i Wiedza (KIW), Warszawa. tyczna [Social and Political Geography], Jacobs, J. (1961), The Death and Life of Great Wyższa Szkoła Nauk Humanistycznych i Dzi- American Cities, New York, Random House. ennikarstwa (WSNHiDZ), Poznań. Jałowiecki, B. (1980), Człowiek w przestrzeni mia- Cambell, B. (1993), Goliath: Britain’s Dangerous sta [Man in Urban Space], Śląski Instytut Places, Virago, London. Naukowy (ŚIN), Katowice. Cichocki, R. and Podemski, K. (1998), Życie Jałowiecki, B. (1988), Społeczne wytwarzanie w Poznaniu 1997. Poznaniacy o swoim mieście przestrzeni [The Social Generation of Space], [ Life in Poznań. Inhabitants on their city], Książka i Wiedza (KiW), Warszawa. Poznańska Drukarnia Naukowa, Poznań. Kotus, J. and Dolata, M. (2004), Społeczne Cichocki, R. and Podemski, K. (1999), Miasto naznaczanie obszarów miasta [Social Mark- w świadomości swoich mieszkańców [The ing of Urban Areas], in Jażdżewska, I. (ed.), City in the Consciousness of its Inhabitants], Konwersatorium Wiedzy o Mieście, Uniwer- Wydawnictwo Fundacji Humaniora, Poznań. sytet Łódzki, pp. 245–256. 22 Michał Dolata and Jacek Kotus

Lefebvre, H. (1991), Production of Space, translat- Poznań Was Not Difficult. On the 19th Cen- ed from French, Blackwell, Oxford, ix, 454p. tury Housing Market], Kronika Miasta Pozna- Lefebvre, H. (1996), Writings on Cities, translated nia, 4: 97–108. from French, Blackwell, Oxford, 250p. Wacquant, L. (2002), The Rise of Advanced Llewellyn, M. (2004), ‘Urban Village’ or ‘White Marginality: Notes on its Nature and Impli- House’: Envisioned Spaces, Experienced cations, in Marcus, P. and Kempen, R. (eds.), Places, and Everyday Life at Kensal House, Of States and Cities, Oxford University Press, London in the 1930s, Environmental and Plan- pp. 221–240. ning D: Society and Space, 22, 2: 229–249. Wallis, A., 1979, Informacja i gwar [Information Martinotti, G. (1996), Four Populations: Human and Noise], Państwowy Instytut Wydawniczy Settlements and Social Morphology in Con- (PIW), Warszawa. temporary Metropolis, European Review, 4, Wellman, B. and Hampton, K. N. (2003), Neigh- 1: 1–21. boring in Netville: How the Internet Supports Mellor, R. (1985), Manchester’s Inner City: the Community and Social Capital in Wired Sub- Management of the Periphery in Newby, in urb, City and Community, 2, Bujra, H., Littlewood, J. and Rees, P. (eds.), 4: 277–311 Restructuring Capital, Macmillan, London. Wiesiołowski, J. (1997), Socjotopografia późno- Mooney, G. (1999), Urban Disorders, in Pile, średniowiecznego Poznania [Sociotopography S., Brook, C. and Mooney,G. (eds.) Un- of Late-Medieval Poznań], Poznańskie To- ruly Cities?, Routledge, London, New York, warzystwo Przyjaciół Nauk (PTPN). pp. 53–102. Wódz, J. (1989), Problemy patologii społecznej Mumford, L. (1937/2002), What is a City, in LeG- w mieście [Problems of Social Pathology in ates, R.T. and Stout, F. (eds.), The City Reader, the City], Państwowe Wydawnictwo Naukowe Routledge, London, New York, pp. 92–97. (PWN), Warszawa. Pain, R. (2000), Place, Social Relations and the Young, L. (2003), The ‘Place’ of Street Children Fear of Crime: a Review, Progress in Human in Kampala, Uganda: Marginalisation, Re- Geography, 24, 3: 365–387. sistance and Acceptance in the Urban En- Pain, R. and Townshend, T. (2002), A Safer City vironment, Environmental and Planning D: for All? Senses of ‘Community Safety’ in New- Society and Space, 21, 5: 607–627. castle upon Tyne, Geoforum, 33, 4: 105–119. Pile, S. (1999), What is a City? in Massey, D., Al- Paper first received: April 2005 len, J. and Pile, S. (eds.) City Worlds, London, In final form: May 2007 Routledge. Reed, Ch. (2003), We’re from Oz: Marking Eth- nic and Sexual Identity in Chicago, Environ- ment and Planning D: Society and Space, 21, 3: 425–440. Robins, K.and Aksoy, A. (1995), Istanbul Ris- ing, European Urban and Regional Studies, 2, 3: 223–35. Robins, K.and Aksoy, A. (1996), Istanbul between Civilisation and Discontent, City 5, 6: 6–33. Sanchez, L. E. (2004), The Global E-rotic Sub- ject, the Ban and the Prostitute-Free Zone: Sex Work and the Theory of Differential Ex- clusion, Environmental and Planning D: Soci- ety and Space, 22, 6: 861–884. Skuratowicz, J. (1999), Budowanie w Poznaniu nie było sprawą trudną. O XIX wiecznym rynku mieszkań [Housing Construction in FROM MIGRATION TO SEGREGATION IN THE FORMER CLOSED CITY

MICHAEL GENTILE Stockholm School of Economics Box 6501, SE-11383 Stockholm, Sweden E-mail: [email protected]

Abstract: Based on the case of the military-industrial city of Ust’-Kamenogorsk, Kazakhstan, this paper explores (a) Soviet and post-Soviet era migration into former closed cities, and (b) the present housing situation of migrant groups living in them, paying particular attention to their ethnic background. The study is based on a survey carried out by the author and the regional statistical authority in January 2001. The principal findings suggest that there has been a clear increase in migrants from the oblast’s rural areas to the regional capital, which is attributable to the regional urbanisation pressure which had been created during the city’s period of ‘closure’, and that the origin of these migrants has shifted in favour of areas with larger Kazakh popula- tions. Also, contradicting the Soviet goals, and resulting from structural factors re-enforced by the closed city regime, the ethnic housing gap is greatest among those who arrived during the Soviet period.

Key words: Closed cities, Kazakhstan, migration, ethnicity, housing, residential segregation.

INTRODUCTION no-compositional aspect of the migration flows will be covered extensively. The fo- The paper studies (a) Soviet and post-So- cus on housing which follows is justified by viet era migration into former closed cities, the findings from previous research which and (b) the housing situation of the migrant suggest that migrant-fed industrialization groups living in such cities in light of their in the non-Russian republics favoured the time of arrival. The example of Ust’-Ka- migrant population—most of which came menogorsk (Kazakh Öskemen, population from Russia—with regard to the housing approx. 300,000) will be taken as a case in allocation system. This happened often point, based on survey data collected by the at the expense of the titular ethnic group author and the oblast’ statistical authority of (Lubin 1984; Tammaru 2001b; Kährik East Kazakhstan in January 2001. 2002; Gentile and Tammaru 2006; Blejere The analysis moves through various et al. 2005). In non-Russian closed cities, scales, but the emphasis will fall on the where the process of industrialization was hitherto poorly studied intra-regional level. strictly regulated and accomplished in con- Furthermore, given the dramatic turna- certed development with the Soviet military round in the ethnic migration patterns in effort, the relative advantage of the indus- the Former Soviet Union (FSU), the eth- trially skilled core nation(s) of the Soviet 24 Michael Gentile

Union, the Slavs, was presumably even more the depopulation of the far northern regions evident.1 By the time of the demise of the and the ‘sunbelt’ effect in the far southern Soviet Union, we may surmise, the titular regions of Russia have accelerated (Den- ethnic groups of many former Soviet repub- isenko and Khorev 1996; Heleniak 1997; lics were underprovided in terms of hous- 1999), whereas a reversal in the previously ing and basic urban amenities vis-à-vis the negative population trend of rural areas has Soviet-era immigrants. With the transition been identified, with pre-urban rural areas from central planning to market allocation, absorbing the most significant of population and the growth of the service sector and increases (Wegren et al. 1997). On the other relative shrinkage of heavy industry, the hand, there has been a dramatic change in structural preconditions of the dynamics migration flows between the Russian Fed- of Soviet ethnic segregation are nowadays eration and the countries nowadays termed absent. However, the inherited legacies in as the ‘Near Abroad’, particularly the Baltic the form of Soviet ‘socialist’ socio-spatial States, Kazakhstan, Kyrgyzstan and Moldo- and ethno-spatial patterns are not (Gentile va. After decades of Slavic in-migration, 2003b). In this paper, we compare an aspect these regions have experienced substantial of legacy—Soviet urban in-migration pat- net losses among the Slavic (Zayonchko- terns—with an aspect of transition—post- vskaya 1999) and German (Brown 2005) Soviet urban in-migration patterns. As we components of their populations. Cou- will see, the two differ considerably, but is pled with a distinct tendency towards de- post-Soviet in-migration eroding the socio- creased fertility, this has resulted in a total spatialities of the previous industry-fed mi- population decrease of notable magnitude gration? This is the field which is explored (Masanov 2002, p. 18; Sadovskaya 1999, in this paper, and the particular case of p. 119). Conversely, the population decrease the former closed city—and we should not in the Russian Federation, largely a result of forget that it was a rather common occur- increased mortality rates (Mäkinen 2000), rence—is expected to yield illuminative has been somewhat compensated by a wave results. of return migration which is approaching ex- The structure of the paper is as follows. haustion (Becker et al. 2003, p. 229; Zayon- First, the migration context of the FSU will be chkovskaya 2007). International migration of outlined and the areas needing more research this type is the focus of interest of a sizeable put into evidence. Then, existent research on volume of literature (for example, Dunlop Soviet migration management and its signifi- 1994; Vitkovskaya 1999; Zayonchkovskaya cance will be presented, from which the pa- 1999). Recently, this interest has expanded per’s prime research questions will be extrapo- to encompass ‘non-traditional’ inmigration lated. Finally, the data, methods and results of (i.e., refugees from Afghanistan, illegal and the study will be reported. legal labour migrants from East Asia, etc.) and its accommodation within immigration policy (Popson and Ruble 2000). In contrast, THE POST-SOVIET MIGRATION CONTEXT inter-regional flows have been followed to a lesser extent (for the Soviet period, Ball The demise of the Soviet system has had and Demko 1978; Mitchneck 1991; Cole and a radical impact on migration in the regions Filatotchev 1992; for the post-Soviet period, of the ex-USSR. Late Soviet trends, such as Wegren et al. 1997; Heleniak 1997; Wegren and Drury 2001),2 whereas intra-regional 1 In Russia, as elsewhere in the USSR, the develop- (intra-oblast’) flows are thus far virtually ment of the cities which hosted significant military-in- dustrial activities took place at the expense of the city’s 2 non-defence related activities. In the absence of any par- Nevertheless, there has been a somewhat surpris- ticularly clear ethnic division of labour, this resulted in ing upsurge in interest for migration-related issues con- unequal access to urban amenities of unequal quality by cerning the Russian Far North and East (Heleniak 1999; socio-professional status (see Gaddy 1996). Kontorovich 2000; Thompson 2004; Round 2005). From Migration to Segregation… 25 unstudied.3 Part of the explanation is the of a broader discussion on the geodemogra- lack of detailed data beyond oblast’-level, phy of the USSR and of its regions for plan- which implies that other sources and meth- ning purposes (for example, Akhmedova et ods are necessary if one wishes to bring the al. 1976). Within this group, there also is analysis down to the local level. growing literature on the various forms of Put differently, we know little about the forced migration which took place and are actual volumes of intra-regional migration, still occurring within the Soviet (e.g. Polyan nor do we know much about the socio-com- 1999) and post-Soviet realms (e.g. Zayon- positional characteristics of the migrant chkovskaya 1997; Pilkington 1998; Riddle groups, the specificities of their principal and Buckley 1998). The second group focus- destinations, and the causalities of their ex- es on explaining the observed patterns with istence. Furthermore, because of the scar- a primary focus on the determinants of the city of the data, there is not much in-depth migration event (for an overview, see Kulu research of a comparative nature discuss- 2003 and 2004). The third group concerns ing migration before and after the collapse the factors which diverted or prevented mi- of the Soviet Union. This paper addresses gration, or which were at least expected to both issues within the unstudied context of divert/prevent it. Rather than emphasising the former closed city. The choice of this the pull factors at the potential migrants’ particular case is motivated by the fact that place of origin, this body of literature recog- the group of former closed cities was by no nises that, under conditions of central plan- means a small one, given the secrecy require- ning, population movements must necessar- ments of an essentially militarised economy ily be managed and planned to ensure the (Gaddy 1996). Likewise, previous research optimal (i.e. most plan-coherent) territorial shows that the closed city’s urbanization distribution of labour, the adequate pres- context under central planning was remark- ervation of state secrets, and the desire for ably different from that of cities which were urban development in line with the concept not subjected to particular regimes of clo- of the ‘socialist’ city (Gang and Stuart 1999, sure (Gentile 2003a; 2004b). p. 118). Such planning and managing of migration made use of economic, adminis- trative and propagandistic (‘educational’) MIGRATION, MIGRATION MANAGEMENT methods (Kumskova 1983, p. 80). AND CLOSED CITIES The Soviet authorities attempted to curtail unplanned or undesired migration The migration literature on the Soviet Union by means of various sorts of administra- can roughly be divided into three inter-relat- tive restrictions, with the so-called internal ed groups. The first group, which is domi- passport and propiska (registration) system nated by works of Soviet vintage, includes at the forefront position (Lewis and Row- studies whose primary focus is to provide an land 1979; Morton 1984; Matthews 1993; accurate description of migration patterns Buckley 1995; Popov 1995a-b; Gaddy 1996; and flows in the region (e.g. Moiseenko 1997; Höjdestrand 2004; Wegren and Drury 2001). Kendirbaeva 1997), often within the context To a varying extent, this strategy inspired most other socialist polities, for example 3 There is, however, the notable exception of “pen- Albania (Sjöberg 1992), Romania (Ronnås dulum” migration (commuting). Commuting has been 1982) and the People’s Republic of China particularly targeted by Soviet scientists since the late 1970’s (see for example Orlova and Sitnik 1981; Kenig- (see Buckley 1995, p. 912). Agreement as to shtein 1983; Khorev and Likhoded 1983). Further- whether such measures actually worked has more, eased by the recent data improvements, there is yet to be reached. Buckley (1995) contends a growing and by now rather voluminous literature on suburbanisation, particularly on the urban regions of the that the success of the limits imposed on the Baltics (Tammaru 2001a; Tammaru 2005.; Leetmaa and growth of specific cities, which were presum- Tammaru 2007; Krišjāne et al. 2005). ably to be enforced through the propiska 26 Michael Gentile system, was at best slight. However, she em- certain concern which the literature rarely phasises that the propiska gained in impor- acknowledges, and usually at best en passant tance after the unplanned migrant’s arrival to (for example, in Wegren and Drury 2001, the city, being the base for access to all sorts p. 40, it is relegated to an endnote). Within of important urban amenities, particularly the group of closed cities, the degrees of clo- to public sector housing, but also to schools, sure surely varied greatly from case to case hospitals etc. On the other hand, based on and at different points in time, in response sample of 308 cities, Gang and Stuart (1999) to the development and perceived secrecy show that urban growth restrictions indeed requirements of the powerful military-in- did matter, and that unrestricted cities grew dustrial complex (Gentile 2004b). Even so, about twice as fast as restricted ones. Fur- we may surmise that in-migration to such thermore, they identified significant growth closed cities must have been subject to much differences between regions and, based stricter control than in-migration to open cit- on a smaller but more secure sample, be- ies, including the ones subject to restrictions tween cities with total expansion restrictions on in-migration or industrial expansion. (i.e. propiska-enforced) and restrictions on It may also be assumed that most closed cit- industrial expansion (and therefore of the ies do not coincide with the totally restricted demand for labour). Surprisingly, the limits ones reported by Lewis and Rowland (1979), on industrial expansion seemed to produce Buckley (1995) or Gang and Stuart (1999), stronger results than the total restrictions, which casts some doubt on the accuracy of implying that economic means were more these scholars’ conclusions regarding the effective in curtailing migration than the efficacy of administrative restrictions on in- direct forms of administrative control of the migration, as the better sealed cities seem to propiska type. This point allows for some have fallen under the open category in their reconciliation with Buckley’s (1995) more analyses. sceptical view of the efficacy of the propiska, Given the military’s important 20–30% but it also indicates that the requirements of share of the Soviet GDP (Kotkin 2001, economic planning (in labour input terms, in p. 61; Åslund 2003), the group of closed cit- this case) preceded the social and political ies might indeed be more important than goals of the system of administrative migra- hitherto imagined. Certainly, this does not tion control. mean that 20–30% of the Soviet population The problem is that, as mentioned above, actually lived in closed cities—after all, the restrictions were also imposed on cities military-industrial complex was well repre- which involved significant military activities, sented in such metropolises as Moscow and to the extent that some were even kept en- Leningrad—but a recent study by Gentile tirely secret, even though the latter’s share (2004b) suggests that, in 1989, up to almost of the total population was relatively small 70% of the total population of Kazakhstan’s (see Rowland 1999). The picture is further 19 largest cities, excluding the two capitals complicated by the fact that many other cit- of Almaty (until 1997) and Astana (since ies—and it is impossible to know which ones 1998), might have resided in an urban place or how many4—were effectively closed, not to which access was strictly limited. just to migrants without a propiska, but to Recent studies on urbanization in the all unauthorised visitors. This is an issue of Former Soviet Union (FSU) indicate that former closed cities and their umlant differ 4 This problem stems partly from the fact that the from their open counterparts with regard concept of the closed city is used rather lightly in the to their population development since the literature. Some authors use it as synonym for ‘totally restricted city’, i.e., a city subject to a strict propiska re- demise of the USSR and the administrative gime. Such a view would mean that most large cities in restrictions on urban-bound in-migration the USSR were ‘closed’, and that Moscow still is, at least associated with it (Gentile 2003a; 2004b). on paper (see Hoffmann, 2002). Importantly, these studies suggest that From Migration to Segregation… 27 former closed cities were, on the one hand, tional migrants throughout the 1990s, and more developed from an infrastructural that many of them settled in the peri-urban point of view, with a better supply of hous- rural areas, at least in part due to the urban ing, schools and other necessities. On the housing shortage. Conversely, during the other hand, access to closed cities was re- same period Kazakhstan was a net sender— stricted, and in-migration was certainly in the year 2001 the in-migration to rural tightly controlled and selective, favouring areas from abroad was almost three times skilled migrants from afar at the expense of lower than was its counter-flow, and in East the relatively untapped regional labour sup- Kazakhstan almost six times lower, i.e. only ply. Accordingly, closed city gates were as- 527 persons (Agenstvo 2002c). The delayed sumed to limit potential urbanization from underurbanization character of the settle- local sources, causing a mounting regional ment system surrounding former closed cit- urbanization pressure during the city’s pe- ies in Kazakhstan is mainly fuelled by inter- riod of closure. The abolishment of most nal migration, which is in contrast with the cities’ regime of closure by the early 1990’s Russian experience described in Wegren enabled the sudden unleashing of this pres- et al. (1997; see also Wegren and Drury 2001, sure. Consequently, in Kazakhstan, the p. 24). population development of former closed Summing up, the findings reported above cities turned out to be less negative than that suggest that former closed cities may be ex- of open ones. Furthermore, the population pected to reveal significantly different long- size of the settlements within commuting term migration trajectories than their open distance of closed cities also experienced counterparts, especially in pre-/post-Soviet a relatively positive development during the comparisons. Given the number of closed 1990’s, displaying a pattern reminiscent of cities in the post-Soviet urban network and the one predicted by the underurbanization the suggested influence exerted by the previ- thesis which was simply labelled ‘delayed un- ous regime of closure on the current region- derurbanization’ (Gentile 2003a; for more al migration trends, it is proposed that this on the underurbanization thesis see Kon- category be analysed more carefully to the rad and Szelényi 1977; Murray and Szelényi extent that this is possible given the limited 1984). Under both delayed and ‘regular’ tools available for research. The theoretical underurbanisation, the settlement pattern contribution of such a study lies in that it ad- arises from a shortage of urban housing, but vances our understanding of the effects of the cause of this shortage differs. During the the more extreme forms of migration regu- post-Soviet epoch, and especially in light lation which were commonplace under cer- of the large number of vacancies resulting tain circumstances (prevalence of the mili- from mass out-migration, the shortage aris- tary-industrial complex) in the urban areas es from the population’s and, especially, the of the Soviet-type polities. Furthermore, an migrants’ low purchasing power, rather than adequate understanding of migration proc- from limited supply. Compared to single- esses at the intra-regional level is necessary family dwellings, living in Soviet-era apart- for various levels of public administration in ment blocks implies relatively high mainte- order to make informed planning decisions. nance costs which cannot be influenced by In light of the above discussion, the fol- the individual (electricity, gas, heating etc.). lowing research questions are set forth: Within the Russian context, Wegren et al. • Where did the urban-bound migrants (1997) identified similar quasi-underur- come from during the ‘closed’ period, and banizational tendencies of rural population where did they come from during the first increase in the areas surrounding the cities ‘open’ decade? of the European non-Chernozem region. • What was the ethnic composition of However, this similarity conceals the fact the closed period migrants, and what has it that Russia was a net receiver of interna- been during the open period? Figure 1. Map showing the location, administrative subdivision and major cities of the East Kazakhstan oblast’. From Migration to Segregation… 29

• What is the housing situation of Soviet tus with regard to the web of administrative era migrants as compared to that of the post- restrictions imposed on population move- Soviet migrants? ment and in-migration to cities. Ust’-Ka- • Are there any differences between the menogorsk was closed, Semipalatinsk open.5 housing situations of the Kazakh migrants Ust’Kamenogorsk has paved sidewalks, vis-à-vis the Russian migrants? Semipalatinsk, for the most, does not. Ust’- To answer these questions, in-migra- Kamenogorsk has tramlines, Semipalatinsk tion to the former closed city of Ust’-Kame- does not. Ust’-Kamenogorsk has various nogorsk (pop. approx. 300,000), East Kaza- technologically advanced Soviet-era pro- khstan oblast’, will be analyzed using survey duction facilities, Semipalatinsk has a sym- material collected in early 2001. Samples of bolically powerful post-2000 suspension persons who had immigrated to Ust’-Kame- bridge. In many ways, Ust’-Kamenogorsk is nogorsk during and after the Soviet period what Semipalatinsk is not, which is why they have been extracted from the larger sample are useful opposites for comparison when of 1,836 respondents, which also includes the possible. native population. Unfortunately, though, Ust’-Kamenogorsk is inhabited by a Rus- the sample does not include persons who sian ethnic majority, even though its share had immigrated to Ust’-Kamenogorsk but of Kazakhs has increased during the 1990s have emigrated since. Likewise, the sample (Alekseenko, 1994; survey)(Table 1). In con- of migrants excludes those who have been trast, the former open city of Semipalatinsk, involved in an intra-urban mobility event the city of Dostoyevskii’s exile, is much more since their first arrival. Therefore, our anal- Kazakhified, with the share of Kazakhs ysis is limited to those who have migrated to in its population having increased signifi- Ust’-Kamenogorsk and remained there ever cantly—from 26.5 percent in 1989 to over since without ever moving within the city. 50 percent in the year 2002.6 The increased Despite the ensuing skewness in the sample, share of Kazakhs in Semipalatinsk is most the findings will be distinct enough to allow likely due to natural increase, combined a number of interesting conclusions. with stronger Slavic out-migration, rather Before moving on to the details of the than to Kazakh in-migration. In Semipal- data and methods employed, a few words on atinsk, the rate of natural growth in 2002 the case study city and its region are war- was of 0.89 ‰, whereas there was a -3.39 ranted (Figure 1). Located in the north- ‰ natural decline in Ust’-Kamenogorsk eastern corner of Kazakhstan, the East Ka- (Upravlenie, 2003a; 2003b). Unfortunately zakhstan oblast’ contains two similarly sized there are no data available for rates of natu- cities which both used to be oblast’ capitals ral population change within cities broken until 1997: Ust’-Kamenogorsk and Semi- by ethnicity, but figures at the national level palatinsk (Semey). Comparing the fate of reveal that Kazakhs have a high annual rate these two cities is instructive. Ust’-Kameno- gorsk evolved into an important node in the 5 Of course, the territory of the Semipalatinsk Test- Soviet military-industrial complex, host- ing Polygon was entirely off limits, but it was located at over 100 km from Semipalatinsk proper. In fact, it hosted ing a number of non-ferrous metallurgical a closed (and secret) city of its own, Semipalatinsk-126, enterprises of all-Union significance, with nowadays Kurchatov. close or direct ties to the military-industrial 6 The figures for 2002, however, refer most likely to the territory of the city administration rather than to complex. Semipalatinsk became the host of a that of the city per se. If this is the case, it is worth con- large food-processing industry, particularly sidering that the city administration territory includes for meat products. Owing to these differ- a dispersed rural population of approximately 30,000 in- dividuals (Agenstvo 1999, p. 41; Agenstvo 2002a, p. 87; ent production specialisations and relative Agenstvo 2002b), most of whom Kazakhs. Therefore, priority status during the Soviet epoch, the it would be reasonable to revise the share of Kazakhs to, population development of these two cities possibly, 40% of the city’s population, whereas Russians differed considerably, and so did their sta- would probably reach approximately 50%. 30 Michael Gentile of natural growth, whereas Russians are ex- last’-level migration balance of the group (ap- periencing sharp natural decline (and out- proximately -28,000 individuals) (Alekseenko migration)(Agenstvo, 2001, p. 58), which 1994). Hence, given moderate Russian natu- may help explain the rapid Kazakhification ral growth, the city’s relative closure and the of Semipalatinsk. oblast’s Russian rural population loss (which

Table 1. Share of Russians and Kazakhs in the population of the major cities of East Kazakhstan oblast’ in 1989 and 2001 (†Zyryanovsk rayon [1989] or city administration [2002]; ‡as of the year 2002)

Russians Russians Kazakhs Kazakhs City (1989, %) (2001, %) (1989, %) (2001, %)

Ust’-Kamenogorsk 81.5 74.2 10.6 18.6

Semipalatinsk 59.6 40‡ 26.5 51.1‡

Source: survey; Agenstvo, 2000a; Alekseenko, 1994, p. 16; Upravlenie statistiki VK oblasti (2003a; 2003b); Goskomstat SSSR (1991, p. 21).

Ust’-Kamenogorsk experienced popula- does not seem to have been equally compen- tion growth throughout the Soviet period, sated by urban growth), the Russian popula- mainly due to positive net migration. Between tion growth of Ust’-Kamenogorsk must to 1979 and 1989, the city grew by 17.4%, with a certain degree have been the result of in- 64.5% of this growth attributable to the Rus- migration from outside the oblast’, most likely sian ethnic group. During the same period, from Russia, though certainly not to the same the city’s Kazakh population grew at a rate extent as in the first two post-war decades. six times higher than that of the Russians. Ust’-Kamenogorsk experienced a four- At the oblast’ level,7 Kazakhs increased 4.2 percent population decline during the times faster than Russians, presumably be- 1989–1999 inter-census period, which is cause of their high but decreasing birth rate a rather moderate decrease compared to (Alekseenko 1994). This suggests that Ka- the standards of the time—neighbouring zakh in-migration to Ust’-Kamenogorsk dur- Leninogorsk and Zyryanovsk, for instance, ing the 1980s might stand for about one third shed 18 and 17% of their populations, re- of the increase (i.e., about 5,000 individuals) spectively (Agenstvo 2000b). However, since in the size of the city’s Kazakh 1989 popula- 1999, the city’s population decline has con- tion of about 34,000. Russians, on the other tinued steadily. The out-migration of Slavs hand, increased largely as a result of in-mi- to the other republics of the CIS is the main gration, although it is not known from where. explanation, together with a modest (0.3%) In the meantime, the oblast’ was experiencing annual natural population decline (Upravle- net migration losses, especially of Russians, nie 2003b). whose numbers decreased by over 30% in As is the case in most of the FSU, the several southern rayony (Kurchum, Tarbaga- urban economy of Ust’-Kamenogorsk was tay, Katon-Karagay and Markakol’ [now part severely damaged by the economic turmoil of Kurchum]). These ethnic Russian losses during the crisis of transition from social- did not ‘reappear’ commensurately in any of ism to a market economy in the early 1990’s. the other rayony, because of the negative ob- Since 1997, however, industrial output has started to pick up again, and the enterprises 7 At the time, the East Kazakhstan oblast’ was only started to re-hire (Moskal’tseva 09/2000). half its current size, as Semipalatinsk oblast’, abolished The demand created by the relatively high in 1997, had not yet been incorporated in it. wages within heavy industry injected new From Migration to Segregation… 31 energy in Ust’-Kamenogorsk’s commercial Before the distribution of the questionnaires and service functions. Accordingly, the city forms, around the 20th of December 2000, became the main magnet for migrants in the the survey was presented in the principal oblast’, which partly happened at the expense oblast’ media (regional television, radio and of similarly-sized Semipalatinsk, whose in- newspaper). The questionnaire forms were dustrial sector remains largely inactive.8 subsequently distributed and collected dur- ing the evening hours and on week-ends by 58 professional statisticians (45 in Ust’- DATA AND METHOD: THE CITIES Kamenogorsk), who acted as interviewers. OF THE RUDNYI ALTAY DATABASE In Ust’-Kamenogorsk, they were rewarded with 50 KZT (approximately US$ 0.30 at the The Cities of the Rudnyi Altay database col- time), and, as a result, only 2% declined par- lects information on the socio-economic sta- ticipation. An additional 5.1% of the dwell- tus of the three cities of Ust’-Kamenogorsk, ings in the sample had been abandoned, Leninogorsk and Zyryanovsk, based on an evacuated or demolished since the register extensive questionnaire survey (N=3,136 was created, which gave a response rate of adult respondents, >18 years of age) car- almost 93%.10 ried out in January 2001 by the author in The sample covered 1,836 respondents in cooperation with the statistical authority of Ust’-Kamenogorsk, 700 in Leninogorsk and East Kazakhstan oblast’, based in Ust’-Ka- 600 in Zyryanovsk, and was extracted from menogorsk, and with the participation of the the complete register of households created city statistical offices in Leninogorsk and on occasion of the 1999 census of the Repub- Zyryanovsk. The overall goal of the survey lic of Kazakhstan. Of the 1,836 respondents was to create a knowledge base on the so- in Ust’-Kamenogorsk, 1,516 belong to an cio-geographical characteristics of small and original sample, and 320 were added to about medium-sized post-Soviet industrial cities 15 neighbourhoods which had few respond- undergoing transition. ents in the original sample. This implies that The author pre-tested the questionnaire the extended sample is somewhat less repre- form, which was in Russian, in a minor pi- sentative than the original one. However, it lot study in December 2000 (N=64) and is the author’s opinion that the benefits ob- supervised the entire subsequent survey tained from increased sample size outweigh process, mostly from the premises of the the risks incurred from allowing a certain statistical authority in Ust’-Kamenogorsk.9 bias towards certain neighbourhoods. This is because the data used for this paper target 8 Even though the available pool of migrants mostly the population operationally defined as mi- consists of ethnic Kazakhs from the rural areas of the grant, i.e., those whose previous place of res- oblast’, it is interesting to note that Semipalatinsk, with a considerably larger share of ethnic Kazakhs in its popu- idence was beyond the municipal boundaries lation, does not appear to be a particularly attractive of the city in which they resided at the time destination. Employment opportunities—or a percep- of the survey. This rather narrow definition tion of their existence—are more important migration motivators than ethnic composition in independent implies that individuals that might have in- Kazakhstan. Previous research shows that this might migrated to the city are excluded from the be otherwise for Russians and other ethnic minorities migrant sample whenever they have been in- (IOM 1998; Sadovskaya 1999; Becker et al. 2005). 9 The questionnaire form was divided into four sec- volved in an intra-urban mobility event since tions (themes) and 41 questions: (a) General informa- tion about the respondent, (b) Housing status, (c) Hous- 10 In line with common local practice, refusals and ing and living conditions and (d) Employment. The topic abandoned or demolished dwellings were replaced by of migration is covered en passant in all four themes, neighbours living in the dwelling immediately to the with the exception of the last few questions which ex- right (a similar strategy was adopted in a UN-funded plicitly concern potential out-migration, making them survey-based study by O’Hara et al., 2007). A detailed less relevant for this paper. The questionnaire form, and discussion of the method problems encountered while a translation into English, may be obtained from the au- working with the survey is provided elsewhere (Gentile, thor upon request. 2004a). 32 Michael Gentile their arrival to the city. This depends on the We will begin by presenting the general fact that the database only registers the most differences in the patterns of in-migration to recent mobility event, whether it be intra-ur- the three cities before and after the demise of ban or external. the Soviet Union at the aggregate level. After For clear reasons, the survey also excludes taking the example of Ust’-Kamenogorsk to those who had migrated to the city during pre- elucidate the ethnic dimension of Soviet and vious years but have left since, either to Rus- post-Soviet in-migration, the section will pro- sia or elsewhere. This means that this article ceed by examining in-migration to the city in essentially compares those persons who had greater geographical detail, emphasising the moved to Ust’-Kamenogorsk before 2001, origin of the migrants and the changes that are still alive and living there, and have never have occurred in this respect since 1991. Final- moved within the city since their arrival. In a ly, with the ethnic variable in mind, the housing Western context, such limitations would have situation of ‘Soviet’ and ‘post-Soviet’ migrants decimated the size of the initial sample; in the will be compared in order to establish whether low mobility contexts of socialism and post- there are any significant differences. socialist transition, the situation is different. Summing up, the sample certainly un- derestimates the size of the migrant group, GENERAL PATTERNS OF IN-MIGRATION even within the rather low mobility context. TO UST’-KAMENOGORSK AND THE RUDNYI Perhaps paradoxically, this is particularly ALTAY REGION CITIES true for the ‘post-Soviet’ migrant sample (mi- grants who have arrived during the 1990s), as Table 2, which is based on the entire Ust’-Ka- its respondents arrived after the introduction menogorsk sample, shows that two thirds of of market relations in the provision of hous- the population of the city were born outside ing, thus making their position more volatile. of its administrative boundaries. This stands Nevertheless, the size and quality of the mi- in contrast with the cases of neighbouring grant sample should allow the extrapolation Zyryanovsk and Leninogorsk, where only of valuable insights for Soviet/post-Soviet about one third of the inhabitants were born comparison. In the following sections, the outside of their respective cities, suggesting migrant sample will generally be divided into that Leninogorsk and Zyryanovsk have long those who arrived during the Soviet epoch, stopped absorbing significant volumes of and those who arrived during the post-Soviet in-migrants. Specifically, 38 percent of Ust’- period until 2001. For the sake of simplicity, Kamenogorsk’s citizens were born within the these sub-samples will be referred to as ‘So- oblast’, primarily in rural areas, whereas 22.2 viet’ and ‘post-Soviet’ migrants, but the read- percent were born outside of Kazakhstan. er should remain aware that the operational A modest 6.3 percent were born in a differ- definition used is rather narrow. ent Kazakhstani oblast’, approximately half of which in urban areas. The situation is quite different in RESULTS: MIGRATION AND SEGREGATION Leninogorsk and Zyryanovsk. First of all, IN UST’-KAMENOGORSK11 locals prevail by birth.12 Secondly, the share of those born abroad is somewhat greater This empirical section will proceed by de- than that of the population born in a differ- scribing and analysing the migrant sample ent locality of the oblast’. As in Ust’-Kame- (based on the unconventional operational nogorsk, few originate from the other oblasti definition provided above) extracted from of the republic. Therefore, contradicting two the Cities of the Rudnyi Altay 2001 database. of Ravenstein’s classical migration ‘laws’

11 This section is based on the survey unless indi- 12 In this case, ‘regional’ is a very relative term, for cated otherwise. the size of the oblast’ is comparable to that of Poland. From Migration to Segregation… 33

Table 2. Place of birth of respondents born outside of Ust’-Kamenogorsk’s official boundaries and their share in the city’s total population, based on entire sample of 1,836 respondents

Urban Rural % of total Place of birth (N) (N) Total population

Other rayon within oblast’ 115 583 697 38.0

Other oblast’ 59 57 116 6.3

Abroad (mostly Russia) — — 408 22.2

Total born outside of city — — 1,221 66.5

Source: survey.

(see Ravenstein 1885, pp. 198–199, and jority of those who arrived during the Soviet Boyle et al. 1998, p. 60), that most migration period are Russian (75.5%), whereas almost is short-distance and of step-by-step char- half of the post-Soviet migrant sample, on acter, the share of international migrants the other hand, is Kazakh (46.2%), although in Leninogorsk and Zyryanovsk is greater Russians still do stand for a significant share than that of ‘regional’ migrants, which is (43.7%). The share of other ethnicities has surprising given the cities’ relatively small declined slightly, from 11.8% to about ten sizes. However, Ravenstein’s observations percent. However, it should be noted that are based on the assumption that migration the ethnic composition of the group of other flows are voluntary. This was not necessarily ethnicities has changed. The pre-1991 sam- the case in the Soviet Union, where coerced ple mostly includes Ukrainians and Ger- (e.g. prison labour) and semi-voluntary mans, whereas post-1991 sample includes (e.g. specific job assignments after gradua- many Tatars and Chechens. tion from higher education) migration was The share of Russians in the urban- implemented when the centrally determined bound migrant sample is perhaps not as in- national and local economic and political teresting as that of the Kazakhs. With the goals were not achievable otherwise. majority of the rural population within the oblast’ being of Kazakh ethnicity, the Kaza- khs’ 12.7% share of the pre-1991 Ust’-Kame- ETHNIC DIMENSION OF SOVIET nogorsk migrant group is very low indeed, AND POST-SOVIET IN-MIGRATION suggesting that the city’s gates were not open TO UST’-KAMENOGORSK to all. More evidence of this can be found in Table 4, which shows that there were con- The ethnic composition of the migrant siderably more Russian in-migrants from sample in Ust’-Kamenogorsk has changed within the oblast’ than Kazakhs (see below), since 1991 (Table 3). An overwhelming ma- suggesting that the gates were not only open

Table 3. Ethnic composition of migrants to Ust’-Kamenogorsk during the Soviet and post-Soviet periods

Kazakhs Russians Other Total Year of arrival (N) % (N) % (N) % (N) %

Before 1991 26 12.7 154 75.5 24 11.8 204 100

1991–2001 110 46.2 104 43.7 24 10.1 238 100 Source: survey. 34 Michael Gentile restrictively, but also ethnically selective. 1995, ch. 6), implying that the volume of in- Somehow, the Kazakhs were underurban- migration to Ust’-Kamenogorsk should have ised (see Tammaru 2001b), and precluded been significantly smaller during the 1980s from the city privileges. This does not neces- anyway. sarily imply that there were direct forms of In the 1980s, the bulk of intra-oblast’ ethnic discrimination at work, although the migrants was Russian, a pattern which was possibility cannot be excluded. Instead, the reversed during the 1990s. The inter-ob- economic policy of intense industrialization last’ and international migrant group was inevitably favoured the Slavic ethnic groups, dominated by ethnic Russians during both whose skills and educational background decades, but the share of inter-oblast’ Ka- better met the demands of the industrial en- zakh migrants has increased considerably. terprises in Ust’-Kamenogorsk. This pattern It is interesting to notice that the in-migra- is not an unknown one: the link between tion of Russians from within the oblast’ economic and ethnic policy has already sur- during the 1990s seems to be the continu- faced elsewhere and was particularly evident ation of a trend established during the in the Baltic States (Park 1994; Gentile and 1980s, if not earlier, whereas Kazakh in- Tammaru 2006). migration seems to be experiencing an ex- A comparison between the origin and plosion. This makes it easy to suggest that ethnicity of the migrant sample to Ust’-Ka- the Kazakhs were in a disfavoured position menogorsk during the 1980s and 1990s un- when applying for an urban residential per- covers additional interesting geographical mit, and that the closed city of Ust’-Kame- features (Table 4). At a first glance, it may nogorsk was more closed for Kazakhs than seem that there was much more migration for Russians. However, such a conclusion to Ust’-Kamenogorsk during the 1990s than needs to be treated carefully, for the dis- during the previous decade. Although the crepancy could depend on spurious ele- magnitude of this gap is probably consider- ments, for instance divergent educational ably smaller than it appears to be because levels between ethnic groups and, there- of the characteristics of the migrant sample fore, different ethnic employment patterns used in the study, many of those who arrived (which may, of course, have been the result during the 1980s have probably already of other discriminatory practices)(Lubin moved at least once within the city. Even 1984; Kaiser 1995). In this light, closed so, because of the absence of a legal hous- cities exacerbated the underurbanization ing market, Soviet cities were characterized of the titular ethnic group (see Tammaru by relatively low residential mobility (French 2001b).

Table 4. Ethnicity and origin of migrants to Ust’-Kamenogorsk between 1981 and 1991 and between 1991 and 2001

Migrants 1981–1991 Migrants 1991–2001 Migrants to (N) (N) Ust’-Kamenogorsk from: Russian Kazakh Other Russian Kazakh Other

Within oblast’ 32 19 2 70 97 10

Other oblast’ 81214126 Other republic 14 0 3 20 1 8

Total 54 20 7 104 110 24

Source: survey. From Migration to Segregation… 35

GEOGRAPHICAL ANALYSIS OF IN- move to a more desirable location, such as MIGRATION TO UST’-KAMENOGORSK the ‘sun belt’ of the southern part of the Eu- ropean USSR. Let us proceed to the detailed analysis of the The most significant changes in the pat- origin of migrants to Ust’-Kamenogorsk dur- tern of in-migration to Ust’-Kamenogorsk ing the Soviet period as a whole (those who occurred during the post-Soviet period arrived before 1991), its last decade (those (1991–2001). Not only has the volume of in- who arrived between 1981 and 1991), and migration (size of the sample) increased con- the post-Soviet period (those who arrived siderably (with the reservation that compari- between 1991 and 2001)(Table 5). The data son is complicated by the characteristics of concerning the former Semipalatinsk oblast’, the migrant sample used in the study), but its which was abolished in 1997, are included in geographical sources have changed. Recent the figures for the East Kazakhstan oblast’. in-migrants typically come from the Kaza- There were three types of Soviet-era kh-dominated, rural, and relatively sparsely intra-oblast’ (regional) migration flows di- populated Katon-Karagay, Kurchum and rected towards Ust’-Kamenogorsk. The first Tarbagatay rayony, together with Zyry- flow, accounting for nearly one third of all anovsk and Ulanskoe rayony. The increase regional migrants, is from the other major in migrant arrivals from Ulanskoe rayon is cities of the oblast’, the second flow (over surprising and might have an ethnic dimen- one third of the regional migrants) is from sion because (unlike the case of Glubokoe the neighbouring rayony, and the third flow rayon), a significant share of the rayon’s is from the more distant Kurchum rayon. population consists of Kazakhs. Again, this Inter-oblast’ (inter-regional) flows came al- suggests that Kazakhs might have been pre- most exclusively from the city of Almaty and cluded from the local urbanisation process from Almaty oblast’, with the addition of of the Ust’-Kamenogorsk region during the a little group originating from South (Yuzh- Soviet epoch. This suspicion is strengthened nyi) Kazakhstan oblast’. International mi- by the fact that other rayony, with even larg- grants stood for a significant share of all mi- er Kazakh population shares, experienced grant arrivals to Ust’-Kamenogorsk. similar in-migration jumps, whereas ray- Similar patterns prevailed in the 1980s, ony with a prevailingly Russian population although the bulk of the migrants from the (Glubokoe, Shemonaikha, Leninogorsk) city of Almaty had already arrived by 1981. experienced only moderate increases since It also appears that the regional and inter-re- the 1980s. One possible explanation lies in gional flows had become relatively more sig- the well-documented discrepancy between nificant at the expense of the international the quality of education in rural areas and flow. This probably reflects the increasing the qualification requirements for industrial labour shortage in the European USSR, jobs in the Soviet Union (see Wegren et al. which no longer acted as prime labour sup- 1997). In a centrally planned quasi-labour plier (Ball and Demko 1978), as well as the market (Mitchneck and Plane 1995), this shift in migration motives—partly facilitated implies limitations on rural-urban migration by the toning down of the orgnabor (organ- supported by coercive measures. But then ized recruitment of labour) system—from again, these limitations should apply to the productivist to environmentalist. In other predominantly Russian rural areas as well. words, migration no longer strictly echoed Certainly, part of the explanation is that the employment and career opportunities a segment of the Slavic population is attract- available. The latter tendency is confirmed ed by competing international destinations, by Mitchneck and Plane (1995) in a study particularly (according to the survey) Altay- that shows that migrants in late-Soviet Rus- skii kray, Tomskaya oblast’ and Novosibir- sia were prone to accept less-paid and lower- skaya oblast’ of the Russian Federation, but prestige jobs if this meant that they could the likelihood of trans-boundary migration 36 Michael Gentile

Table 5. Previous place of residence of respondents whose previous dwelling was located outside of Ust’- Kamenogorsk

Migrants to Ust’-Kamenogorsk (N)

Year of arrival

Origin: –91 81–91 91–

Other rayon within oblast’ 115 55 177

Semipalatinsk city administration 13 5 12 Zyryanovsk 12 8 19 Ust’-Kamenogorsk city -- - administration Leninogorsk city administration 10 5 6 Zaysan 4 2 9 Ayagoz 2 2 4 Glubokoe 22 9 14 Zharma 3 2 0 Katon-Karagay 6 4 23 Kokpekty 2 1 9 Kurchum 14 7 22 Tarbagatay 3 2 18 Ulanskoe 12 3 29 Urdzhar 2 0 0 Shemonaikha 9 5 11 Beskaragay 1 0 0 Borodulikha 0 0 0

Other oblast’ 19 9 32

Akmolinsk and Astana (city) 1 1 2 Aktiubinsk 0 0 0 Almaty (city) 5 1 9 Almaty 4 4 10 Atyrau 1 1 0 Zhambyl 1 1 2 Zap.-Kazakhstan 1 0 0 Karaganda 1 1 5 Kustanay 1 0 0 Kzyl-Orda 1 0 1 Mangistau 0 0 1 Pavlodar 0 0 1 Sev.-Kazakhstan 0 0 0 Yuzh.-Kazakhstan 3 0 1

Abroad (mostly Russia) 69 17 29

Source: survey. From Migration to Segregation… 37 straight from rural areas is not very high, exodus towards Russia and other post-So- mainly because of the high economic and viet republics. Part of this return migration social costs of migration. may consist of skilled metallurgical workers Therefore, it is beyond any doubt that returning to Ust’-Kamenogorsk in response the oblast’s rural Slavs were already be- to resumed production, for example at the ing urbanised during the Soviet era (and Titanium-Magnesium Combine (Shaya- for various reasons at the expense of the khmetov 13/09/2000). Kazakhs)(Alekseenko 1994; survey). Hence, at the eve of the collapse of the iron curtain, and of the strict permit regime in Ust’-Ka- THE HOUSING SITUATION OF MIGRANTS menogorsk, a large group of Kazakhs was IN UST’-KAMENOGORSK waiting to be urbanised and was allowed to do so upon the opening of the city’s gates in The data presented above suggest that the early 1990s. It is also worth noting that a strong inflow of migrants to the former the former oblast’ boundary between East closed city and current oblast’ capital of Kazakhstan and Semipalatinsk oblasti is re- Ust’Kamenogorsk has taken place since the flected in both Soviet and post-Soviet migra- USSR’s demise. We now proceed to see how tion figures. In-migration from the former this flow is being accommodated by the city’s Semipalatinsk oblast’s Ayagoz, Zharma, housing supply. Beskaragay and Borodulikha rayony was and We will seek to establish whether there remains low or inexistent. One would expect are any significant differences between the Semipalatinsk to attract migrants from these general housing situation of migrants who rayony, but other studies suggest that this city arrived during the Soviet and post-Soviet does not attract as many migrants as Ust’-Ka- periods. The assessments are based on four menogorsk (Gentile 2003a; 2004b). This sug- types of variables—dwelling type, the level gests that Semipalatinsk (which was open in of the respondents’ satisfaction with the Soviet times) had not experienced the same dwelling, the availability of three common urbanization pressure build-up during the amenities (cold running water, connection Soviet period, and/or that economic reasons to municipal sewerage, and telephone line), for migration prevail (there are fewer—and and living space measured as the number less paid—jobs in Semipalatinsk than in Ust’- of persons per room (excluding kitchen and Kamenogorsk). Ust’-Kamenogorsk’s relative bathroom) and surface per capita in square prosperity may well be attributable to the metres. The initial analysis will be followed longue durée of the city’s high priority status by a comparative assessment of the role of under central planning. In short, the findings ethnicity before and after 1991. on intra-regional migration indicate that Ust’- The data presented in Table 6 indicate Kamenogorsk’s regime of closure was partic- that the post-Soviet migrants are more likely ularly strict towards the ethnic Kazakhs, who to live in apartments than the Soviet mi- remained ‘underurbanized’ throughout the grants, whereas the latter are about twice as Soviet period. Conversely, the open city of likely to live in detached houses. Further- Semipalatinsk was indeed open to all. more, about 18% of the post-Soviet migrants Inter-regional migration has increased live in a communal apartment or rented considerably since the 1980s, though it con- room (both generally considered inferior tinues to originate primarily from Almaty in terms of residential utility) rather than in city and Almaty oblast’, with the addition a dwelling of their own, a share which is of Karaganda oblast’, whereas internation- three times higher than that for the Soviet al in-migration has increased moderately migrants. As dwellings, apartments are (although net international migration is, of generally more comfortable than detached course, negative), probably reflecting the re- houses (only about 1% of the latter have run- turn migration generated by the early 1990s ning water, see Gentile and Tammaru 2006), 38 Michael Gentile but they do not allow the household to earn pect, post-Soviet migrants indeed do live in extra income (or save) through backyard poorer housing, it is necessary to proceed by agriculture. However, there are clear dif- analyzing the ‘objective’ indicators. ferences within the two dwelling types, and Dwellings occupied by post-Soviet mi- general conclusions cannot be drawn on the grants are slightly more likely to have internal basis of this distinction. access to cold running water and are clearly Table 6 also suggests that the representa- more likely to be connected to the municipal tives of the post-Soviet migrant sample are sewerage system, which only has marginal only marginally less satisfied with their cur- significance, for the majority of the detached rent hou si ng c ond it ion s . However, a s Sz elény i housing stock is not connected to the sewer- (1983, p. 141) points out, people tend to be age system, even though many houses are realistic about their housing chances, opting quite well-equipped otherwise. Post-Soviet to answer that they are relatively satisfied migrants are also slightly more likely to have to general questions about housing quality. a private telephone line reaching their dwell- For example, a recent survey in Poland sug- ing. On the basis of these three indicators, gests that Poles have rather modest housing it would seem as though post-Soviet migrants aspirations (Kaltenberg-Kwiatkowska 2002, live in slightly better equipped dwellings. p. 116). If this is true and if, as one may sus- In a certain sense, this is not surprising. If it is

Table 6. Housing characteristics of Soviet and post-Soviet migrants in Ust’-Kamenogorsk

Soviet migrants Post-Soviet migrants

N%N%

Dwelling type Apartment 104 51.0 144 60.3 House 88 43.1 51 21.3

Communal apartment 12 5.9 31 13.0

Room 0 0.0 12 5.0

Level of satisfaction Very satisfied 14 6.9 14 5.9 Satisfied 111 54.4 113 47.3

Not very satisfied 35 17.2 54 22.6

Unsatisfied 44 21.6 57 23.8

% of dwellings inhabited % of dwellings inhabited by Soviet migrants by post-Soviet migrants

Amenities Cold running water 79.9 81.1 Sewerage connection 54.9 74.8

Telephone 35.3 38.1

Dwellings inhabited Dwellings inhabited by Soviet migrants by post-Soviet migrants

value % of city value % of city average average

Living space Persons per room 1.15 98.3 1.64 140.2 m2 per capita 23.35 126.6 18.58 100.7

Source: survey. From Migration to Segregation… 39 true that (mainly Russian) urban-bound im- migrants, although some insights about the migrants in the non-Russian republics were aggregated group of ‘other’ ethnicities will be granted privileged access to housing in the provided. The data indicate that both Soviet form of units in new prefabricated apartment and post-Soviet Russian migrants are more blocks with ‘full conveniences’ (Ruoppila likely to live in apartments and less likely and Kährik 2003), as the Soviet parlance has to live in communal apartments or rented it, then these dwellings should be over-rep- rooms (Table 7). Nonetheless, housing sat- resented among the units that were vacated isfaction among Russians and Kazakhs is during the exodus of the early 1990s, which relatively similar. Kazakh post-Soviet mi- mainly involved ethnic Russians. grants seem to be somewhat more satisfied The data concerning the level of crowd- and Russian post-Soviet migrants somewhat ing in housing suggest quite a different story. more dissatisfied with their housing situation Whereas Soviet migrants live at a density of than their Soviet’ counterparts. Is this a sign 1.15 persons per room, there are 1.64 per- that previous Russian ethnic privileges are sons per room in housing occupied by post- gradually starting to dissipate? The remain- Soviet migrants. In relative terms, these der of the data that are going to be discussed figures are equal to 98.3 and 140.2% of the provides a surprising though somewhat con- city average. Hence, in terms of housing, it tradictory picture. appears that Soviet migrants are even bet- Soviet migrants generally live consider- ter off than the average inhabitant of Ust’- ably more spaciously, but in slightly less mod- Kamenogorsk, which is perhaps part of the ern dwellings. However, both Soviet and, es- explanation as to why they still live in their pecially, post-Soviet Russian migrants live in current dwelling rather than having moved. better-equipped dwellings than their Kazakh This might reflect a more generous alloca- counterparts. Other ethnicities seem to be tion at the outset, or it could mean that somewhere in between. In terms of crowding, number of household members has changed the disparities are even greater. Kazakh Sovi- through either a decrease (death, children et migrants are worst off, with less than 15m2 moving out, etc.) or an increase (childbirth, of living space per capita, i.e. just 80.1% of relatives moving in, subtenants, etc.). Post- the city average. By contrast, Russian Soviet Soviet migrants, on the other hand, live in migrants have over 25m2 per capita (138.4% much more cramped conditions than the of the city average), which is rather high even average citizen. In order to exclude the pos- by the standards of the European USSR, and sibility that the post-Soviet migrants might other ethnic Soviet migrants have 18.57m2 live in housing with larger rooms—which per capita. In terms of persons per room, would be the likely case in the coveted Sta- the pattern is replicated: Kazakh Soviet mi- lin-era apartments—per capita living space, grants live at a density of 1.48 persons per including kitchen and bathroom, has been room (126.5% of the city average), as com- calculated as well. The previous picture is pared to 1.08 for Russians (92.3%) and 1.26 confirmed and reinforced: Soviet migrants for other ethnicities (107.6%). Given that the live considerably more spaciously, with state-sector built dwellings were largely built 23.35 square metres per capita (126.6% of by the metallurgical enterprises and allocat- the city average of 18.45m2), as compared to ed to their own staff and their families, an 18.58m2 (100.7%) for post-Soviet migrants. overwhelmingly Slavic population segment To put it shortly, post-Soviet migrants live (Gentile 2005), we may conclude that Kaza- in somewhat better quality dwellings but in khs that did make it through the gates of the much more cramped conditions than their closed city during the Soviet period probably Soviet counterparts. did so in order to work in the low priority Let us now proceed to the ethnic di- economy (i.e. in schools, hospitals, light in- mension. The focus will primarily be on the dustry, etc.), which was not able to meet the distinctions between Russian and Kazakh housing demands of those employed by it. 40 Michael Gentile

Table 7. Housing characteristics of Soviet and post-Soviet migrants in Ust’-Kamenogorsk by ethnicity

Soviet migrants Post-Soviet migrants

Kazakhs Russians Other Kazakhs Russians Other (N) (N) (N) (N) (N) (N)

Dwelling type Apartment 9 83 12 52 77 15

House 9 67 12 27 16 8

Communal 8402371 apartment Room000840

Level of Very satisfied 2 111563 satisfaction Satisfied 10 88 13 53 48 12 Not very 7 25 3 22 25 7 satisfied Unsatisfied 7 30 7 30 25 2

% of dwellings inhabited % of dwellings inhabited by Soviet migrants by post-Soviet migrants

Kazakhs Russians Other Kazakhs Russians Other

Amenities Cold running 73.1 76.0 62.5 75.5 89.4 70.8 water Sewerage 61.5 54.5 50.0 69.1 82.7 66.7 connection Telephone 26.9 37.0 33.3 30.9 42.3 50.0

Dwellings inhabited Dwellings inhabited by Soviet migrants by post-Soviet migrants

Kazakhs Russians Other Kazakhs Russians Other value % of city average value % of city average value % of city average value % of city average value % of city average value % of city average Living Persons per 1.48 126.5 1.08 92.3 1.26 107.6 1.86 159.0 1.41 120.5 1.58 135.0 space room m2 per capita 14.78 80.1 25.54 138.4 18.57 100.6 16.3 88.3 20.5 111.0 20.76 112.5

Source: survey.

The housing stock occupied by the post- losses in living space, when compared to the Soviet migrants also conceals a gap between Russian Soviet migrants. This indicates that Russians and Kazakhs in terms of per capita the Russians have lost their formal privileges living space and crowding. The Kazakh post- in access to housing. The housing gap which Soviet migrants, with their 16.3m2 average, remains between the Russian and Kazakh are somewhat better off than Kazakh Soviet post-Soviet migrants is most likely due to two migrants in terms of living space, but they major factors. On the one hand, at the city are also worse off than the Russian post-So- aggregate level, the Russians have a higher viet migrants, who still occupy over 20m2 per disposable income than the Kazakhs—by capita. On the other hand, the Russian post- almost 20%, according to the survey. On the Soviet migrants have experienced substantial other hand, it is beyond any doubt that the From Migration to Segregation… 41

Kazakhs are discriminated against in the Gaddy (1996, p. 154)—parasitized the rest housing rental market. A quick look at the of the economy in the sense that ‘there was local private advertisement newspaper con- no trickle down effect from the defence in- firms this,13 as there are plenty of advertise- dustry to the rest of the region’ and ‘on the ments which explicitly state that the dwelling whole, the development and well-being of in question may only be rented out to ‘Eu- the defence industry was at the expense of ropean’ nationalities or that ‘Asians’ are not the rest of the local economy’. welcome. As long as this attitude proceeds In fact, if Ust’-Kamenogorsk may be tak- and most dwellings remain occupied by ‘Eu- en as a sufficiently representative example, ropeans’, the Kazakhs will retain their dis- we may conclude that the housing stock of advantaged position in the rental market. closed cities was (and is) somewhat more po- The findings reported above support larised than that of open cities,14 and that the the idea that Russians were very strongly enterprise-based housing allocation system favoured in the allocation of housing during contributed to the ethnification of this di- the Soviet years, and that housing allocation vide. The housing divide, in turn, is the tan- by need (rather than by merit, presuming gible foundation of socio-economic and eth- that Russian workers were generally more nic residential segregation (Gentile 2005). skilled) was a concept that did not manage to Nowadays, segregation is supported by the move from ideology to practice. This ethnic increasing income gap on the one hand, and favouritism occurred in an indirect manner by a certain degree of ethnic discrimination, through the industrial enterprises’ housing on the other. In the past, Soviet policies in- construction and allocation functions, which directly favoured industrial workers, espe- ensured preferential (de facto exclusive) ac- cially within the military industrial complex, cess to housing for the workers. The latter and therefore Russians. Again, as both Ru- were predominantly Russian, especially ble (1989) and Tammaru (2001b) state, eco- within the industrial enterprises which en- nomic policy becomes ethnic policy. joyed highest priority during the Soviet peri- od (Gentile 2005; see Kaiser 1995, for more on the representation of the titular ethnic MAIN CONCLUSIONS groups in Soviet industry and other branches of the economy). Such enterprises were also This paper has contributed to migra- the ones most enveloped by security regu- tion and segregation research on post-Soviet lations, including the closure of the city in countries by comparatively analysing data on which they were located, and were forced to urban-bound migrants towards the former offer a housing ‘carrot’ in exchange. But this closed city of Ust’-Kamenogorsk. Despite carrot was not available to everyone—on the shortcomings of the sample, three irrefu- the contrary, it only concerned the work- table trends have been brought to the fore. ers engaged in the most sensitive activities. • First of all, in-migration to Ust’-Kame- The rest of the population lived as usual, in nogorsk has changed dramatically. The most a city with poor housing resources, greater significant development is the clear increase environmental problems, but somewhat bet- in migrants from the oblast’s rural areas to ter sidewalks and public transportation. The the regional capital, and the origin of these defence industry—as previously noted by migrants has shifted in favour of the rayony with a larger Kazakh population, suggesting 13 Although the author currently does not have ac- that particular segments of the population, cess to such publications, an example taken from the website of the Iz ruk v ruki (10/12/2003) private adver- 14 tisement newspaper in Almaty may help illustrate the With the exception of the most extreme cases of situation: ‘Room in one-room apartment, Gogol str., 3rd closure, where it is presumed that all families were al- floor, for rent to working European lady, 8000 tenge/ located self-contained apartments with generous floor month’. space in exchange for not existing on the map. 42 Michael Gentile particularly rural Kazakhs, were closed off enabled them to parasitize and damage the from the regional urbanization process dur- rest of the urban economy, and the Kazakh ing the Soviet period. Restrictions mattered population disproportionately so. Gradually, to some. the effects of these past legacies will be erod- • Secondly, and as a consequence of the ed by the realities of transition. For the time above, the ethnic composition of the former being, however, the Soviet-made Slavic eth- closed city-bound migrant flow has shifted in nic advantage in housing, as indeed in many favour of the previously poorly represented other spheres of urban life in the northern Kazakhs. Together with the out-migration of half of Kazakhstan, still persists. Elsewhere, Slavs and Germans, this is gradually chang- such as in the Baltic States, it has long since ing the ethnic structure of the city. dissolved, but there the relative advantage • Thirdly, in terms of housing provision, was marginal at the outset. The situation those who arrived during the Soviet period is thus quite different in Central Asia, and are generally better off than the post-So- particularly in the region’s former closed cit- viet migrants. However, within both groups, ies, where the ethnic consequences of Soviet there is a sizeable ethnic gap: the Russians industrialisation run deeper. are considerably better-housed. In contrast with the official Soviet goals of equality, the ethnic gap within the Soviet migrant group ACKNOWLEDGEMENTS: is larger than that within the post-Soviet migrant group. This reflects the housing dif- The author wishes to thank the Swedish Re- ferences indirectly created as a result of the search Council (Vetenskapsrådet) and the relative priority ascribed to certain indus- Wallander-Hedelius foundation of Svenska trial enterprises, coupled with the structural Handelsbanken for generous financial sup- factors which hindered the Kazakhs from port, and Örjan Sjöberg (Stockholm School being at equal terms with the Russians in the of Economics), Charlotta Hedberg (Uppsala labour market and, accordingly, the housing University), Jan Öhman (Uppsala Univer- allocation system. sity), and two anonymous referees for their The findings of this paper shed some new valuable comments on previous drafts of this light on the relation between migration and paper. segregation in the Soviet Union, and par- ticularly in the closed city. It is reaffirmed that ethnic segregation in the USSR resulted REFERENCES from the past practice of blending indus- trial policy with housing policy. In order to Agenstvo Respubliki Kazakhstan po statistike achieve the centrally established economic (1999), Kratkie itogi perepisi naseleniya v Re- targets, the high priority enterprises were spublike Kazakhstan 1999 [Short Results of forced to offer a bundle of benefits in order the Population Census of the Republic of to attract and retain their personnel, a prac- Kazakhstan], Almaty, Agenstvo Respubliki tice which occurred at the expense of those Kazakhstan po statistike. employed in the low priority sectors of the Agenstvo Respubliki Kazakhstan po statis- economy. The low-high priority distinction tike (2000a), Natsional’nyi sostav naselenia is also a well-acclaimed ethnic one (Lubin Respubliki Kazakhstan—2 tom [Nationali- 1984; Sacks 1992; Tammaru 2001b; Gentile ties in the Republic of Kazakhstan—vol. 2], 2005; Gentile and Tammaru 2006), and its Almaty, Agenstvo Respubliki Kazakhstan po implications at the level of unequal access to statistike. housing have been shown to be all the more Agenstvo Respubliki Kazakhstan po statistike significant in the context of the closed city, (2000b), Chislennost’ i Razmeshchenie Nase- where the power and prestige of the high lenia Respubliki Kazakhstan tom 1, [Size and priority industrial actors was such that it Distribution of the Population of the Republic From Migration to Segregation… 43

of Kazakhstan – vol.1], Almaty, Agenstvo Responses of Women and Men during Eco- Respubliki Kazakhstan po statistike. nomic Turmoil: Lessons from Kazakhstan, Agenstvo Respubliki Kazakhstan po statistike Eurasian Geography and Economics, 44, (2001), Demograficheskii Ezhegodnik Kaza- 3: 228–243. khstana 2000, [Demographic Yearbook of Becker, C., Musabek, E, Seitenova, A. and Kazakhstan 2000], Almaty, Agenstvo Re- Urzhumova, D. (2005), The Migration spubliki Kazakhstan po statistike. Response to Economic Shock: Lessons from Agenstvo Respubliki Kazakhstan po statistike Kazakhstan, Journal of Comparative Eco- (2002a), Chislennost’ naselenia gorodov Re- nomics, 33, 1: 107–132. spubliki Kazakhstan (na daty provedenia Blejere, D., Butulis, I., Zunda, A., Stranga, perepisey sootvestvuyushchikh let) [Urban A. and. Feldmanis, I. (2005), Istoriia Latvii Population of the Republic of Kazakhstan XX vek [History of Latvia in the 20century], (as reported by the respective Censuses] Riga: Jumava. Boyle, P., Halfacree, K. and Robinson, V. (1998), (13/02/2002). Exploring Contemporary Migration, Harlow Agenstvo Respubliki Kazakhstan po statistike (UK), Longman. (2002b), Chislennost’ naselenia Respubliki Brown, A. (2005), The Germans of Germany Kazakhstan po oblastyam, gorodam, rayonam, and the Germans of Kazakhstan: A Eurasian rayonnym tsentram i poselkam na 1 yanvarya Volk in the Twilight of Diaspora, Europe-Asia 2002 g. [Population of the Republic of Kaza- Studies, 57, 4: 625–634. khstan by oblast, urban settlement, rayon, Buckley, C. (1995), The Myth of Managed Migra- rayon centres and rural settlements, Janu- tion: Migration Control and Market in the So- ary 1, 2002], Electronic database purchased viet Period, Slavic Review, 54, 4: 896–916. from the Statistical Agency of the Republic of Cole, J., and Filatotchev, I. (1992), Some Obser- Kazakhstan . vations on Migration within and from the Agenstvo Respubliki Kazakhstan po statistike Former USSR in the 1990s, Post-Soviet Geog- (2002c), Migratsia Naselenia Respubliki raphy, 33, 7: 432–453. Kazakhstan za yanvar’–dekabr’ 2001 g [Mi- Denisenko, M. and Khorev, B. (1996), Migratsiya gration of the Population of the Republic Naseleniya v Rossii v Nachale 90-kh Godov of Kazakhstan, January–December 2001], i Eë Perspektivy [Population Migration in Electronic database purchased from the Sta- Russia at the Beginning of the 1990s], Izvesti- tistical Agency of the Republic of Kazakhstan ia Russkogo Geograficheskogo Obshchestva . 128, 6: 13–18. Akhmedova, K.B., Dvoskin, B.Ya. and Semeno- Dunlop, J. (1994), Will the Russians Return from va, M.I. (1976), Proizvoditel’nye Sily Vostoch- the Near Abroad?, Post-Soviet Geography and nogo Kazakhstana [The Factors of Production Economics 35, 4: 204–215. in Eastern Kazakhstan], Alma-Ata, Nauka. French, A. (1995), Plans, Pragmatism and Peo- Alekseenko, A. N. (1994), Narody Vostoch- ple—The Legacy of Soviet Planning for Today’s nogo Kazakhstana [The Nations of Eastern Cities, London, UCL Press. Kazakhstan], Ust’-Kamenogorsk: VKO Dom Gaddy, C. (1996), The Price of the Past—Russia’s Druzhby narodov, VK oblastnaya adminis- Struggle with the Legacy of a Militarized Econ- tratsia and Vostochnyi Gumanitarnyi Insti- omy, Washington DC, Brookings Institution tut. Press. Åslund, A. (2003), Building Capitalism, New Gang, I. and R. Stuart (1999), Mobility, where York, Cambridge University Press. Mobility is Illegal: Internal Migration and Ball, B. and G. Demko (1978), Internal Migration City Growth in the Soviet Union, Journal of in the Soviet Union, Economic Geography, 54, Population Economics, 12: 117–134. 2: 95–114. Gentile, M. (2003a), Delayed Underurbanization Becker, C., Musabek, E. Seitenova, A. and and the Closed-City Effect, Eurasian Geogra- Urzhumova, D. (2003), Short-Term Migration phy and Economics, 44, 2: 144–156. 44 Michael Gentile

Gentile, M. (2003b), Residential Segregation in Kährik, A. (2002), Changing Social Divisions in a Medium-Sized Post-Soviet City: Ust’-Kame- the Housing Market of Tallinn, Estonia, in nogorsk, Kazakhstan, Tijdschrift voor Econo- Housing, Theory and Society, 19: 48–56. mische en Sociale Geografie, 94, 5:. 589–605. Kaiser, R.J. (1995), Nationalizing the Work Gentile, M. (2004a), Divided Post-Soviet Cities? Force. Ethnic Restratification in the Newly Residential Segregation in Leninogorsk and Independent States, Post-Soviet Geography, Zyryanovsk, Kazakhstan, in Geografiska An- 36, 2: 87–111. naler B, 86, 2: 117–136. Kaltenberg-Kwiatkowska, E. (2002), The Hous- Gentile, M. (2004b), Former Closed Cities and ing Situation in Poland: Perceptions of the Urbanisation in the ex-USSR: An Explora- Problem, in Marana, I. and S. Tsenkova (eds.), tion in Kazakhstan, in Europe-Asia Studies, Challenges and Opportunities in Housing: New 56, 2: 263–278. Concepts, Policies and Initiatives, Riga, CIB Gentile, M. (2005), Urban Residential Prefer- and City of Riga, pp. 109–117. ences and Satisfaction in the Former Soviet Kendirbaeva, G. (1997), Migrations in Kaza- Union: Results from a Survey in Ust’-Kame- khstan: Past and Present, Nationalities Pa- nogorsk, Kazakhstan, Urban Geography, 26, pers 25, 4: 741–751. 4: 296–327. Kenigshtein, M.I. (1983), Gorod i Mayatnikovye Gentile, M. and Tammaru, T. (2006), Housing Migranty [The City and the Commuters], and Ethnicity in Ust’-Kamenogorsk, Kaza- EKO, 9, 111: 90–95. khstan, Urban Studies, 43,10: 1757–1778. Khorev, B.S. and Likhoded, V.N. (1983), Rural- Goskomstat SSSR (1991), Sotsial’no-Ekonom- Urban Commuting in the USSR Soviet Geog- icheskoe Razvitie Stolits Soyuznykh i Avtonom- raphy, 24, 8: 570–587. nykh Respublik, Kraevykh i Oblastnykh Tsen- Konrad, G. and Szelényi, I. (1977), Social Conflicts trow [Socio-Economic Development of the of underurbanisation, in Harloe, M. (ed.), Cap- Capital Cities of the Soviet and Autonomous tive Cities, Chichester, Wiley, pp. 157–173. Republics and of the Kray and Oblast Centres] Kontorovich, V. (2000), Can Russia Resettle the Moscow, Informatsionno-Izdatel’skii Tsentr. Far East?, Post-Communist Economies, 12, Heleniak, T. (1997), Internal Migration in Russia 3: 366–383. During the Economic Transition, Post-Soviet Kotkin, S. (2001), Armageddon Averted—The So- Geography and Economics, 38, 2: 81–104. viet Collapse 1970–2000, New York, Oxford Heleniak, T. (1999), Out-Migration and Depopu- University Press. lation of the Russian North during the 1990s, Krišjāne, Z., Bauls, A. and Vilciņš, A. (2005), Post-Soviet Geography and Economics, 40, Changing Patterns of Population Mobility in 3: 155–205. Latvia, Ģeogrāfiski Raksti, XII: 65–73. Hoffmann, D. (2002), The Oligarchs—Wealth and Kulu, H. (2003), Residence and Migration in Power in the New Russia, New York, BBS Pub- Post-War Soviet Estonia: The Case of Rus- lic Affairs. sian-Born Estonians, Tijdschrift voor Econo- Höjdestrand, T. (2004), The Soviet-Russian Produc- mische en Sociale Geografie, 94, 5: 576–588. tion of Homelessness—Propiska, Housing, Pri- Kulu, H. (2004), Determinants of Residence and vatisation, online resource available at http:// Migration in the Soviet Union after World www.anthrobase.com/Txt/H/Hoejdestrand_T_ War 2: the Immigrant Population in Estonia, 01.htm (last accessed 20 June 2007). Environment and Planning A, 36: 305–325. IOM (1998), Emigration Potential of Persons of Kumskova, N.Kh. (1983), Zakonomernosti Ukrainian Origin and Ukrainian Extraction, Dvizhenia Naselenia Kirgizii [Regularities in Residents of the Republic of Kazakhstan, Kyiv, the Population Mobility in Kirgizia] Frunze, International Organization for Migration. Kyrgyzstan. Iz ruk v ruki (10/12/2003), private advertisement Leetmaa, K. and T. Tammaru (2007), Suburbani- newspaper in Almaty, www.izrukvruki.kz/ zation in the Tallinn Metropolitan Area by db/show_p.htms (last accessed 10 December Dwelling Types and Locations of Residence, 2003). Geografiska Annaler, B89, 2: 127–146. From Migration to Segregation… 45

Lewis, R. and Rowland, R. (1979), Population Re- Population], in Valentey, D.I. (ed.), Mayatnik- distribution in the USSR: its Impact on Society, ovaya Migratsia Sel’skogo Naselenia, Moscow, New York, Praeger. Finansy i Statistika, pp. 7–17. Lubin, N. (1984), Labour and Nationality in Soviet Pal’gov, N.N. and Yarmukhamedov, M. Sh. (1970), Central Asia, London, Macmillan. Kazakhstan, Moscow, Mysl. Mäkinen, I. (2000), Eastern European Transition Park, A. (1994), Ethnicity and Independence: and Suicide Mortality, Social Science & Medi- the Case of Estonia, Europe-Asia Studies 46, cine, 51: 1405–1420. 1: 69–87. Masanov, N. (2002), Perceptions of Ethnic and Pilkington, H. (1998), Migration, Displacement All-National Identity in Kazakhstan, in Ma- and Identity in Post-Soviet Russia, London sanov, N., Erlan, K., Chebotarev, A. and Oka, and New York, Routledge. N., The Nationalities Question in Post-Soviet Polyan, P. (1999), Geografia Nasil’stvennykh Mi- Kazakhstan, Chiba (Japan), Japan External gratsii v SSSR [Geography of Compulsory Trade Organization, pp. 1–68. Migrations in the Soviet Union], in Naselenie Matthews, M. (1993), The Passport Society—Con- i Obshchestvo 37 (June). trolling Movement in Russia and the USSR, Popov, V.P. (1995a), Pasportnaya Sistema v SSSR Boulder (CO) and Oxford (UK), Westview. (1932–1976 gg.), chast’ 1 [Passport Regula- Mitchneck, B. (1991), Geographical and Eco- tions in the Soviet Union, 1932–1976, Part I], nomic Determinants of Interregional Migra- Sotsiologicheskie Issledovania, 9: 3–14. tion in the USSR, 1968–1985, Soviet Geogra- Popov, V.P. (1995b), Pasportnaya Sistema v SSSR phy, XXXII, 3: 168–189. (1932–1976 gg, chast’ 2.)[Passport Regula- Mitchneck, B. and Plane, D. (1995), Migra- tions in the Soviet Union, 1932–1976, Part II], tion and the Quasi-labor Market in Russia, in Sotsiologicheskie Issledovania, 9: 3–13. International Regional Science Review, 18, Popson, N. and Ruble, B. (2000), Kyiv’s Nontradi- 3: 267–288. tional Immigrants, Post-Soviet Geography and Moiseenko, V. (1997), Migratsia Naselenia Economics, 41, 5: 365–378. v Perepisyakh Rossii i SSSR [Population Ravenstein, E.G. (1885), The Laws of Migration, Migration in the light of Censuses of Rus- Journal of the Statistical Society, XLVIII, part sia and the Soviet Union], Voprosy Statistiki, 2, June. 3: 30–37. Riddle, L. and Buckley, C. (1998), Forced Mi- Morton, H.W. (1984), The Contemporary Soviet gration and Destination Choice: Armenian City, in Morton, H.W. and Stuart, R.C. (eds.), Forced Settlers and Refugees in the Rus- The Contemporary Soviet City, New York, sian Federation, International Migration, 36, M.E. Sharpe, pp. 3–24. 2: 235–255. Moskal’tseva, Galina Antonovna (06/09/2000), Ronnås, P. (1982), Centrally Planned Urbaniza- former Chief Architect of the city of Ust’-Ka- tion: The Case of Romania, Geografiska An- menogorsk, formal interview. naler B, 64, 2: 143–151. Murray, P. and Szelényi, I. (1984), The City in the Round, J. (2005), Rescaling Russia’s Geogra- Transition to Socialism, International Journal phy: the Challenges of Depopulating the of Urban and Regional Research, 8, 1: 90–107. Northern Periphery, Europe-Asia Studies, 57, O’Hara, S., Hannan, T. and Genina, M. (2007), 5: 705–727. Assessing Access to Safe Water and Moni- Rowland, R.H. (1999), Secret Cities of Russia toring Progress on MDG7 Target 10 (Access and Kazakhstan in 1998, Post-Soviet Geogra- to Safe water and Basic Sanitation), Lessons phy and Economics, 40,4: 281–304. from Kazakhstan, Water Policy (forthcom- Ruble, B. (1989), Ethnicity and Soviet Cities, So- ing). viet Studies, 41, 3: 401–414. Orlova, G. and Sitnik, L. (1981), O Politekonom- Ruoppila, S. and Kährik, A. (2003), Socio-Eco- icheskom Soderzhanii Migratsionnoy Pod- nomic Residential Differentiation in Post- vizhnosti Naselenia [On the Politico-Eco- socialist Tallinn, Journal of Housing and the nomic Bases of the Migratory Mobility of the Built Environment, 18: 49–73. 46 Michael Gentile

Sacks, M. (1992),Work Force Composition, Patriarchy, and Social Change, in Lewis, (31/10/2003). R.A. (ed.), Geographical Perspectives on Upravlenie statistiki VK oblasti (2003b), Gorod Soviet Central Asia, London and New York, Ust’-Kamenogorsk [The City of Ust’-Kame- Routledge, pp. 180–207. nogorsk] at (31/10/2003). i Migratsionnaya Politika v Kazakhstane Vitkovskaya, G. (1999), Potential Migration of [Migration Processes and Migration Policy in Russian-Speaking Populations from Central Kazakhstan], in Zayonchkovskaya, Zh. (ed.), Asia to Russia, in Demko, G., Ioffe, G. and Migratsionnaya situatsia v stranakh SNG, Zayonchkovskaya, Zh. (eds.), Population Un- Moscow, Tsentr Izuchenia Problem Vynuzh- der Duress —The Geodemography of Post-So- dennoy Migratsii v SNG, pp. 119–138. viet Russia, Boulder, Westview Press. Shayakhmetov, Bagdat Mukhametovich (07/09/ Wegren, S., Ioffe, G. and Nefedova, T. (1997), 2000), President of the Ust’-Kamenogorsk Demographic and Migratory Responses to Titanium-Magnesium Combine, formal in- Agrarian Reform in Russia, Journal of Com- terview. munist and Transition Studies, 13, 4: 54–78. Sjöberg, Ö. (1992), Underurbanization and the Wegren, S. and A.C. Drury (2001), Patterns of In- Zero Urban Growth Hypothesis: Diverted ternal Migration during the Russian Transi- Migration in Albania, Geografiska Annaler tion, Journal of Communist Studies and Tran- B74, 1: 3–19. sition Politics, 17, 4, 15–42. Szelényi, I. (1983), Urban Inequalities under State Zayonchkovskaya, Zh. (1997), Vynuzhdennye Mi- Socialism, New York, Oxford University granty iz Stran SNG i Baltii v Rossii [Forced Press. Migrants from the CIS and Baltic States into Tammaru, T. (2001a), Suburban Growth and Sub- Russia], Naselenie i Obshchestvo 18 (April). urbanization under Central Planning: The Zayonchkovskaya, Zh. (1999), Recent Migration Case of Soviet Estonia, Urban Studies, 38, Trends in Russia, in Demko, G.,. Ioffe, G. and 8: 1341–1357. Zayonchkovskaya, Zh. (eds.), Population Un- Tammaru, T. (2001b), The Soviet Union as a De- der Duress—The Geodemography of Post-So- viant Case? Underurbanization in Soviet Es- viet Russia, Boulder: Westview Press. tonia, Urban Geography, 22,: 6: 584–603. Zayonchkovskaya, Zh. (2007), Labour Migration Tammaru, T. (2005), Sub-urbanization, Employ- from Russia. New challenges on the Back- ment Change and Commuting in Tallinn Met- ground of Regulations in the SCHENGEN ropolitan Area, Environment and Planning A, Agreement, paper presented at the Comparing 37, 9: 1669–1687 . Regional Development in the Baltic States in- Thompson, N. (2004), Migration and Resettle- ternational workshop, Riga, 1–3 February. ment in Chukotka: A Research Note, Eura- sian Geography and Economics, 45, 1: 73–81. Paper first received: 2006 Upravlenie statistiki VK oblasti (2003a), Gorod In final form: March 2007 Semipalatinsk [The City of Semipalatinsk] at DEFINING HEAT WAVES—DIFFERENT APPROACHES

MAGDALENA KUCHCIK Insitute of Geography and Spatial Organization, Polish Academy of Sciences, Twarda 51/55; 00818 – Warsaw, Poland E-mail: [email protected]

Abstract: Heat waves have not been defined officially, however there are many non-official defini- tions in use. This paper reviews some of the most frequently used definitions of heat wave and its measures. It also reveals different numbers of heat periods that result from an analysis employing different definitions of a heat wave. This has been exemplified on the basis of data from Poland and confirmed the need for regionally accepted guidelines by which to define heat waves.

Key words: definitions of heat wave, extreme weather, thresholds, synoptic approach, Poland.

INTRODUCTION air temperature was 29°C. The comparison of the air temperature values reached in dif- Heat waves, as extreme weather phenom- ferent events of hot weather indicates the ena, are more dangerous for people than much greater severity of sporadic heat waves any other weather events on account of their occurring in some inland US cities within wide, relatively frequent occurrence, es- continental air mass, as compared with the pecially in highly populated areas, and the costal cities where maritime air prevails great increase in mortality rates they cause. (Table 1). The general perception is that the number In the 1970s, the description of heat of heat waves has increased over the past wave also began to be applied to Europe. decade, and, as the global climate continues At that time, the summer of 1976 was the to warm, the number and intensity of heat hottest summer in the United Kingdom waves and the death tolls they give rise to since records began. As the hot, dry weather might increase considerably. continued the whole summer, an extremely One of the first descriptions of heat severe drought ensued in August, devastat- waves comes from the United States of ing heaths and causing forest fires. Summer America and dates back to the New York of 1994, in turn, brought severe heat in Cen- 1896, in which the weekly mean maximum tral Europe and the Benelux, lasting with Table 1. Examples of heat waves

Place Date Characteristics Effects Authors Comments

New York, USA 1896 mean weekly T>29°C unknown Ellis and Nelson 1978 -

St. Louis, 1936 Mean weekly Tmax up to 39.4°C Rise of mortality in the 40-80 age Ellis and Nelson 1978 Lasted over Missouri, USA group almost 2 months

New York, USA 1948 Daily Tmax up to 38.3°C, mean daily Over 100% rise in M., over 200% Ellis and Nelson 1978 5 days T up to 33.2°C increase among heart and arteries (25-29 August) diseases especially in 60-90 age group

New York, USA 1975 Daily Tmax up to 36.7°C, mean daily 100% rise of deaths from ischemic Ellis and Nelson 1978 7 days T up to 31.1°C heart diseases and strokes (30 July–5 August) especially among 65 and older United Kingdom 1976 Almost one month with Severe drought in August, fires BBC Weather H.W.: 22 June - 16 July; summer

Tmax>26.7°C; within it 1976 - the hottest in UK since

15 following days with Tmax>32.2°C; records began

Central and 1980 Kansas City, T>38°C for 17 1,250 – 10,000 excess deaths, Smoyer et al. 2000 Tmax over 32°C almost every day Southern USA days and T<32°C for 2 days in 3 one of US history’s most from June to September months; Texas, T>48°C devastating natural disasters

Greece 1987 In the suburbs of Athens Tmax up to 1000 excess deaths Matzarakis and Mayer 1991 H.W.: 20-31 July

43.6°C and Tmin>24.4°C Benelux, 1994 Belgium: mean daily T up to Belgium – 1226 excess deaths; Sartor et. al. 1995 H.W.: 22 July – 7 August central Europe 27.5°C; Czech – 294 deaths (11.5% rise in Kyselý 2004

(Czech R., Czech R.: mean Tmax in H.W. total M., 13.6% rise in CVD M.); Kuchcik 2001, 2006 Poland) 33.3°C; Warsaw (Poland) – 132 excess

Poland: mean Tmax in H.W. 33.8°C deaths (33% rise in total M., 37%

in Torun, Tmax up to 37.8°C in in CVD M.) Wroclaw

United Kingdom 1995 Tmax up to 35.2°C (1 August) More than 460 extra deaths (8.9% Rooney et. al. 1998 August 1995, the hottest in UK increase in total M.) since records began (1659).

Chicago, USA 1995 Tmax up to 40°C 700 excess deaths (85% rise in McGeehin, Mirabelli 2001 H. W.: 12 –16 July mortality, 11% rise in hospital Semenza et. al. 1996 admissions) Western and 2003 UK: T=38.5°C; At least 35,000 excess deaths; loss Kosatsky 2005 Many T records, T rose of 20%-

Southern Europe France: Tmax >40°C for 2 weeks of in thickness of glaciers in Alps; Michelozzi et. al. 2004 30% higher than seasonal averages August; drought UNEP 2004 Switzerland: T=41.5° Europe (Poland, 2006 July 2006 - the warmest-ever month 500-1000 extra deaths in Twoja Pogoda [Polish portal Your H.W.: end of June – end of July

Czech R., UK, in most European countries, with Netherlands, at least 370 extra Weather] Many Tmin records Netherlands), monthly mean T higher of 5°C then deaths in USA, severe drought BBC Weather USA, Canada average NOAA

M. – mortality, CVD – cardiovascular, H.W. – heat wave, T – air temperature, Tmax - maximum air temperature, Tmin - minimum air temperature Defining Heat Waves… 49 some breaks from 26th June to 7th August areas of forest by fire. The losses due to Eu- (Table 1). In 2003, western and southern rope’s 2003 heat wave are estimated to have Europe was plagued by a record heat wave exceeded 13 billion euros (UNEP 2004). that extended from northern Spain to the The aim of this paper is to review some Czech Republic and from Germany to Italy. approaches to the definition of the term It produced maxima in the range 35° to 40°C ‘heat wave’ and to show how the number of repeatedly, over almost the whole of Eu- heat waves differ as various definitions are rope. In the United Kingdom, temperature applied. There is thus confirmation of the records were broken with 38.5°C noted on need for a standard definition of the term 10th August at Brogdale in Kent. In France, heat wave on the regional scale at least. temperatures soared to 40°C and remained unusually high for two weeks of August. In Switzerland, June was the hottest month ever TYPES OF ‘HEAT WAVE’ DEFINITIONS recorded in 250 years of archives and a tem- perature record of 41.5°C was established on Paradoxically, the most common way of de- August 11th (UNEP 2004). At least 35,000 fining heat waves is not to define them at all. people died as a result of this heat wave in Many studies, especially the older descrip- France, Germany, Spain, Italy, the UK, the tions of heat waves, and the medical or epi- Netherlands, Portugal, Belgium, Switzer- demiological studies on hot weather-related land and the Czech Republic. France suf- mortality, do not demand the definition of fered the worst losses with more than 14,000 the heat wave (Ellis and Nelson 1978, Roon- deaths, especially among the very old. The ey et al. 1998, Keatinge et al. 2000). They excess mortality in France was estimated at describe unusual hot periods of a few days 20% for the 45–74 years age group, at 70% or weeks, or only separate days with espe- for those aged 75–94, and at 120% for peo- cially high temperatures. And, although the ple over 94 years. Prior to this heat wave only key word ‘heat wave’ appears in such papers, a few deaths during heat waves were declared there is no need to define this precisely. due to hyperthermia, heatstroke or other The term ‘heat wave’ has not gained an classic heat illnesses. However, according official definition from the World Meteoro- to the French medical reports from summer logical Organization, although reports about 2003, 24% of the excess deaths among peo- particularly high and sustained temperatures ple over 74 years of age were directly heat- come from many parts of the world and many related (Kosatsky 2005). definitions are in use. The Oxford English Heat waves take the greatest toll in the Dictionary defines a heat wave as a ‘wave or cities, because Urban Heat Island, which access of excessive heat in the atmosphere’. has been thoroughly examined and docu- The most common meteorological defini- mented, usually intensifies after sunset and tion of a heat wave is a wave of unusually reaches its maximum before sunrise (Oke hot weather which may be accompanied by 1987, Fortuniak et al. 2006, Szymanowski excessive humidity. Although a heat wave is 2005). In rural areas people generally ob- a weather event, it should not be evaluated or tain some relief from the heat at night. The considered without reference to its effects on air pollution which is usually more severe in human beings. Thus ‘the commonly adopted cities than in the countryside can also exac- definition of a heat wave implies that it is an erbate the health-damaging effects through extended period of unusually high atmos- additional stress to the human respiratory phere-related heat stress, which causes tem- and circulatory systems (Katsouyanni et al. porary modifications in lifestyle and which 1993, Larsen 2003). Such extreme weather may have adverse health consequences for conditions also have wide adverse environ- the affected populations’ (Robinson 2001). mental effects on aquatic ecosystems and Such general definitions could be used glaciers, as well as destroy crops and large successfully in describing separate events 50 Magdalena Kuchcik of heat and its impact on the environment perature is in many cases too low to define (fires, droughts), people (morbidity, mortal- the heat, other upper percentiles are in use. ity), or the economy (the costs of the losses). Temperature and time threshold According However, for comparative studies between to the American Meteorological Society in different countries, for any calculations of 1900, A. T. Burrows quite rigidly defined the influence of heat on morbidity or mortal- a ‘hot wave’ as a spell of three or more days ity or for implementing a Heat Health Warn- on each of which the maximum shade tem- ing System coherent for the whole of Europe perature reaches or exceeds 90°F (30.6°C) definition is needed. Also, an impact-related (AMS, Glossary of meteorology). This defi- definition of a ‘heat wave’ must meet the cri- nition with the threshold 30°C is in use in teria that society is susceptible to or unable the Czech Republic (Kyselý 2004). In turn, to cope with these events (Koppe and Jend- the Polish meteorological dictionary defines ritzky 2004, Souch and Grimmond 2004). a heat wave as a period of several days or The majority of the definitions described weeks with maximum air temperature above below are not official, which is to say that 30°C, separated by few cooler periods (Kuch- there is no national agreement for using cik 2001, Niedźwiedź 2003). that definition of the term ‘heat wave’. All The Netherlands Royal Meteorological definitions could be divided into: those us- Institute defined a heat wave as a period of ing thresholds (of separate meteorological at least 5 days, each having a minimum tem- parameters e.g. air temperature or the val- perature of at least 25°C, including at least ues of multivariate indexes) and those using 3 days with a maximum temperature equal synoptic classifications. to or higher than 30°C (Huynen et al. 2001). In Finland, the threshold temperature is THRESHOLDS 25°C. In the United Kingdom a heat wave Thresholds could relate to both single should have an air temperature 3°C higher weather parameters (e.g. air temperature) or than the average for 5 consecutive days. simple indices (e.g. Apparent Temperature) In France, a heat wave is a period of at least and to complex indexes based on human 2 days with daily maximum air temperature heat balance (e.g. Perceived Temperature). above 36°C and a daily minimum air temper- Thresholds are absolute or relative in type. ature above 23°C. In Greece, the threshold An absolute threshold definition of heat for maximum air temperature equals 38°C wave implies that there will be rare heat load for 3 consecutive days (Koppe et al. 2003). in colder regions and frequent heat load in The threshold values naturally grow from warmer regions. Relative thresholds (e.g. the north to south, from 25°C in Finland to 38°C 97% percentile) are based on the assump- in Greece. In regard to the IPCC definition tion that the probability of heat load is of the of the weather event, the upper 10% of the same order of magnitude everywhere. The temperature probability in Athens will be relative thresholds are often based on lo- very different from the same quantile in Hel- cal climate-health relationships and include sinki, and of course heat stress for the organ- the important issue of human adaptation to ism will be greater in Greece than in Finland climatic conditions (Koppe and Jendritzky (Beniston and Stephenson 2004). 2004). According to the Intergovernmental In Canada, a heat wave is defined as Panel on Climate Change (IPCC), heat wave a period of time more than three days long as an extreme weather event is as rare as or of maximum temperatures at or above 32°C. rarer than the 10th or 90th percentile of The Weather Channel in the USA uses the a particular distribution of an atmospheric following criteria for a heat wave: a mini- variable (IPCC 2001). It is evident that the mum of ten states must have 32°C or higher upper 10% of air temperature differs in line temperatures, while temperatures must be at with the baseline climate. However, as the least five degrees above normal in parts of threshold of the 90th percentile of air tem- that area for at least two days. Defining Heat Waves… 51

Climate change researchers use, in ex- or Heat Stress Index that are most common. amining the trends and variations as regards Nevertheless, precise indexes, based on extreme hot periods, the ‘heat wave dura- human heat balance models and demand- tion index’ (HWDI) or ‘warm spell dura- ing more data, are sometimes also applied tion index’. The ‘heat wave duration index’ in defining heat waves; examples being the is defined as the 6 or more consecutive days Perceived Temperature (PT), as calculated during the year over which the daily maxi- using the Klima-Michel-Model (Laschewski mum temperature is in excess of 5°C above and Jendritzky 2002). the normal maximum temperature. The Humidex was first developed in 1965 daily means of maximum temperature are by Canadian meteorologists. It combines calculated for a 5-day window centered on temperature and water vapor pressure into each calendar day from the reference period. one number to reflect the perceived tem- Although extreme hot days were included perature. Humidex remains a useful, and among the numbers, any period of ‘warmer hence popular, means of determining how than normal’ days during the year which met hot one actually feels outside in Canada, the above defined criterion is considered. where its value also generally decreases as In the ‘warm spell duration index’ the maxi- latitude increases (Meteorological Service mum air temperature is replaced by the 90th of Canada). percentile of daily mean temperature (En- vironment Canada …; European Climate Humidex = T + h Assessment…). Therefore, the definitions of h = (0.5555)*(e - 10.0) heat waves in climate-change scenarios on e = 6.11 * exp(5417.7530 * ((1/273.16) - (1/Td))) trend or anomaly maps are not precisely ‘me- where: teorological’ definitions but ‘annual warmer T is dry bulb temperature (°C), e is water va- than normal’ ones. por pressure and Td is the dew point temperature Research on heat wave indicators also (kelvin). explored high night-time temperatures, so- called tropical nights, during which the air The Apparent Temperature (heat index) temperature remains above 20°C even at is probably the best−known and most widely night. High night−time temperatures affect used (USA, many European countries) heat people adversely because of the lag time in stress index. Its basic version derived from the night−time cooling of urban dwellings research by R.G. Steadman (1979), who (Thornbrugh 2001). evaluated human physiological responses Multivariate indexes. Human physical com- to various weather conditions. At the begin- fort is dependent not only on air tempera- ning he took into consideration, not only air ture, but also on the interaction of a number temperature and humidity, but also ventila- of weather variables. For example the hu- tion rate, surface radiation and convection, man response to outdoor conditions is worse clothing resistance to heat and moisture on sultry days than at the same air temper- transport. Later he developed a simplified ature in a dry air mass (Baranowska et al. definition of Apparent Temperature (AT) 1968, Thompson et al. 1996, Changnon et al. using temperature, vapour pressure and wind 2003). speed for conditions indoors and outdoors Over the past several years, numerous (in shaded and sunny locations) (Steadman indices have been developed to measure hu- 1984). As a result of further studies an even man discomfort in heat, these were able to simpler algorithm of AT was developed—us- be used successfully in defining heat waves. ing only two commonly measured inputs of However, among them it is the simplest in- air temperature and dew point temperature dices using widely measured meteorological yet explaining almost all of the variation in elements, such as: Humidex, Apparent Tem- Steadman’s published apparent tempera- perature (heat index), Weather Stress Index ture tables (Steadman 1979, Kalkstein and 52 Magdalena Kuchcik

Valimont 1986, Michelozzi et al. 2000). It is to watches or warnings being issued by the (assuming no or a light wind): National Weather Service. Finally, for much of the southern states a heat wave was de-

AT = -2,653 + 0,994T + 0,0153Td fined as a period of at least 48 hours during where: which neither the overnight low nor the day-

T is dry bulb temperature (°C) and Td is the time high heat index falls below 26.7°C and dew point temperature (°C). 40.6°C respectively (Robinson 2001). A heat wave in Italy is defined as the pe- Humidex and AT have predetermined riod over which maximum AT is above the health impacts associated with various levels 90th annual percentile and which begins with of the index (Table 2). an increase of 2°C compared to the previous Threshold levels for both the discussed day (Michelozzi et al. 2004). For calculating indexes could define heat waves. In its hot maximum AT, daily maximum air tempera- weather response plan Toronto Public ture and the dew point temperature from the Health suggests declaring heat warning when same hour as the maximum temperature are there is a predicted or observed Humidex of taken (this not necessarily being the maxi- 40–45°C for at least 2 days, and declaring mum dew point temperature) (Smoyer et al. a heat emergency when the Humidex value 2000). is greater than 45°C (Smoyer and Rainham Apparent Temperature (AT) in its sim- 2001). However in Ontario (Canada), as an plest, exclusively meteorological form is very increase in mortality was noted when the useful in statistical analyses. Like the air maximum heat index exceeded 32°C and this temperature it also gives threshold values level became the indicator of heat-stress days above which the mortality rates are statisti- there (Smoyer 1998, Smoyer et al. 2000). cally significant higher than the baseline rate In the United States, several thresholds (Davis et al. 2003, Michelozzi et al. 2000) for the heat indexes were tested with a view and in some cases it correlates better with

Table 2. Health effects associated with different values for Humidex and Apparent Temperature (AT)

Range (°C) Health impact

Humidex

< 30 No discomfort

30–39 Some discomfort

40–45 Great discomfort; avoid exertion

> 45 Dangerous

> 54 Heat stroke imminent

AT (heat index)

<26.7 Comfort

26.7–32.1 Fatigue possible with prolonged exposure and/or physical activity

32.2–40.6 Sunstroke, heat cramp and heat exhaustion possible with prolonged exposure and/or physical activity 40.7–54.4 Sunstroke, heat cramp, or heat exhaustion likely and heatstroke possible with prolonged exposure and/or physical activity >54.4 Heatstroke/sunstroke highly likely with continued exposure

Source: Environment Canada and the United States National Weather Service (Smoyer-Tomic and Rainham 2001). Defining Heat Waves… 53 mortality than does air temperature only (slight breeze, mean radiant temperature (Kyselý and Huth 2004). equals to air temperature and relative hu- AT served as a basis for two other clima- midity 50%), in which the perception of tological indexes developed to evaluate ge- heat or cold would be the same as under ographical variations in human discomfort the actual conditions. This index takes all during heat waves, and used in the USA: the relevant mechanisms of heat exchange into Weather Stress Index and the Heat Stress account, with due consideration given to Index. The Weather Stress Index (WSI) is well-adapted clothing. Perceived heat and constructed by calculating the AT and com- cold is computed by means of the comfort paring how the AT value on a particular day equation which is based on a complete heat varies from the mean for that day at a par- budget model of the human body and the ticular location. The index ranges from 0 thermophysiological assessment is made to 100%, with the most uncomfortable con- for a male person, the ‘Klima Michel’. ditions at the highest values. This relative Deutscher Wetterdienst issues heat warn- index compares a particular value of appar- ings when strong to extreme heat load is ent temperature to a climatic normal for involved, and there must be 3 or more days a given location, and evaluates how unu- forecast of moderate or higher heat load sual a definite value might be (Kalkstein based on PT (Laschewski and Jendritzky and Valimont 1986). The Heat Stress In- 2002, Koppe and Jendritzky 2004). dex (HSI) is the newest comprehensive summer index to evaluate summer relative SYNOPTIC APPROACH heat stress as the deviation from the norm This approach requires the development (average calculated on the basis of a long of a synoptic or weather type classifica- period). It includes daily minimum and tion that can be more or less sophisticated. maximum AT, mean cloud cover, cooling As an alternative to the ‘individual variable’ degree-days and the number of consecu- approach, the synoptic approach takes into tive days of extreme heat. The statistical account the entire weather situation, rather distribution of meteorological variables than single elements. However, most of the calculated for 10-day periods (from May biometeorological synoptic classifications to September) gives rise to the daily per- are the determination of categories of at- centiles values of each variable, which are mospheric circulation type and the analysis then summed before being fitted to a sta- of weather maps—a prolonged procedure tistical distribution. The daily HSI value that is very often more subjective than ob- at some location is the percentile of each jective. Therefore, despite recent advances summation value. For example, a 97% HSI in the classification of synoptic-scale events, value indicates that only 3% of days on there has been a need to develop a simpler, that date are expected to experience more automated, large-scale air mass-based pro- stressful conditions than that particular cedure. This need was met by the idea of the day. HSI was calculated for each summer Temporal Synoptic Index (TSI) developed day for over 230 weather stations across the by L.S. Kalkstein and his co-workers at the United States over the period 1971–2000, University of Delaware in the 1980s. This is and can easily be used in forecasting condi- widely used in the United States of America, tions that impair human health (Watts and where it describes impacts of hot weather on Kalkstein 2004). mortality comparably or even better than An example of the use in defining heat other methods and has been shown to be suc- waves index, based on a heat budget of the cessful in heat/health studies (Kalkstein and human body is given by Deutscher Wetter- Corrigan 1986, Kalkstein 1991, Kalkstein et dienst in Germany. It uses Perceived Tem- al. 1996a, Kalkstein and Green 1997). perature (PT) in °C, which is defined as the The synoptic procedure was designed air temperature of a reference environment to classify days that are meteorologically 54 Magdalena Kuchcik similar. Under the TSI, each day is defined The result of this evaluation is a calendar in terms of a variety of available meteoro- listing the air mass to which each day has logical variables (such as: air temperature, been assigned. A second discriminant func- dew point temperature, cloud cover, air tion analysis is used to determine whether a pressure, wind speed and direction) meas- day is to be considered transitional between ured at the four main standard times (00, air masses (Kalkstein et al. 1996a, Kalkstein 06, 12, 18 UTC). Then, by way of princi- and Green 1997). pal component analysis (PCA), each day is Types of air masses defined in the SSC expressed in terms of a set of component are: dry polar (DP), moist polar (MP), dry scores. A clustering (average linkage) pro- temperate (DM), moist temperate (MM), cedure is then used to group days with simi- dry tropical (DT), moist tropical (MT), lar component scores into meteorologically moist tropical plus (MT+) and transitional homogenous, distinct weather types (units, (TR). The nomenclature identifies the char- clusters), i.e. types of air mass. Weather acter, rather than source region, of the air map analysis for selected days in each clus- mass. ter allows the general characteristics of The SSC was redeveloped into SSC2, in each TSI air mass to be described. A stand- which the process of the selection of seed ard stepwise multiple regression analysis days was changed into ‘sliding seed days’. then makes it possible to determine which Identification of seed days takes place in factors within the oppressive air mass con- four two-week ‘windows’ throughout the tribute to elevated mortality (Kalkstein year (one for each season), and an algo- and Corrigan 1986, Kalkstein 1991). The rithm to produce a theoretical ‘seed day’ for basic assumptions of the TSI are: that each weather type for each day of the year is the magnitude of the variable should vary created. The four two-week periods shift by significantly between air masses but little location, to correspond with the hottest and within air-masses; that the variable should coldest two weeks and the midway points in be one of the regularly−observed weather between (Sheridan 2002). elements available for many sites; that the Studies funded by the U.S. Environmen- index should be based on surface data only, tal Protection Agency (EPA) and pointing to permit widespread application, with the to very strong relationships between par- most important air mass indicators being ticularly hot and oppressive air masses, those describing heat and moisture proper- allowed for the use of TSI and SSC in im- ties (Kalkstein and Corrigan 1986). plementing systems to warn against heat The Spatial Synoptic Classification waves—the Heat/Health Watch Warning (SSC) was developed in 1990s. It permits in- Systems (HWWS) first applied in 1995 in ter-site comparison of daily types of air mass Philadelphia, and later in other American across a very large region. SSC requires ini- cities, then in Rome and other Italian cit- tial identification of the major air masses, ies, Shanghai and Toronto (Kalkstein et al. on the basis of long series for an extended 1996b, Kalkstein 2004). In most locations it set of meteorological parameters for each is the maritime tropical (or subtropical) air site. ‘Seed days’ representative of the typical mass (MT+) that is most oppressive—with meteorological character of each air mass very warm and humid conditions and high at a location are then selected. Every effort overnight temperatures, as well as dry trop- is made to maximize the number of seed ical (DT)—with hot and dry conditions and days for each air mass (at least 30 being se- high solar radiation. In Rome both of these lected in most cases), without jeopardizing air masses appear on 11–12% of all summer group homogeneity. Seed days are next used days and approximately correspond to days as input for a linear discriminant function that heat waves occur, and are associated analysis to generate a linear function for with 26–35% excess mortality (de Donato each air mass from its group of seed days. et al. 2005, Kalkstein 2004). Defining Heat Waves… 55

AN EXAMPLE OF THE DETECTING OF The third definition is derived from the HEAT WAVES USING TWO DIFFERENT weather service in Toronto, this making it DEFINITIONS clear just how ineffective a definition bor- rowed directly from another city can be, even To show how different the output obtain- if this city is situated in the same (temperate) able using different approaches might be, an climatic zone and in the same type of transi- example of the number and duration of heat tion climate (though south of 9° of latitude). waves in 10 Polish cities over the whole pe- Geographically, the selected cities rep- riod 1993–2002 is to be presented. The heat resent the whole area of Poland (Figure 1). waves were defined on the basis of meteoro- Among them yearly mean air temperature logical data (daily data and those from 00, varies from 7.4°C in Białystok (north-east) 06, 12, 18 UTC) that were available to the to 9.2°C in Wrocław (south-west). The cli- author. Two types of definitions, one based matic conditions change from the milder on air temperature and time threshold and with maritime influences in the west to the the other on multivariate indexes were ap- more severe and continental in the north- plied. Other definitions illustrating the syn- east. The value of the 95th percentile for optic approach could not be tested due to the AT varied geographically in a different way lack of adequate data. The following defini- to air temperature: from 23.2°C in Gdańsk tions were tested: (on the sea, where wind and air humidity • a minimum 6-day period (5 days for mostly reduce the subjective temperature) a heat wave starting in May or June) with to 27.7°C in Rzeszów (in the very south-east, Apparent Temperature (AT) calculated for where climate is more continental than in 12 UTC (2 pm in Poland in a summer) above Wrocław) (Kuchcik and Blazejczyk 2005, the 95th annual percentile and an increase Kuchcik 2006). A daily maximum air tem- of 2°C compared to the previous day. In the perature of 30°C corresponds to the follow- possible 1–day breaks, AT may not drop be- ing percentiles of Tmax: 0.977 in Toruń, 0.978 low the 90th percentile. To define the first in Rzeszów and Wrocław (the warmest city heat waves in a year, which are perceived by in Poland) and 0.996 in Gdańsk (Table 3). humans to be more burdensome, the shorter period was adopted (Kozłowska-Szczęsna et al. 2004, Koppe and Jendritzky 2004). • at least 3 consecutive days with daily maximum air temperature above 30°C—so called tropical days according to the Polish meteorological dictionary (Huynen et al. 2001, Kyselý 2004, Niedźwiedź 2003). • at least 2 consecutive days with Humi- dex of 40°C (Smoyer and Rainham 2001). The first definition was applied for Polish climatic conditions, considering the distribution of air temperature values. The second is in use in, e.g. the Czech Republic and The Netherlands. Maximum air tem- perature and air temperature from 12 UTC (that is used in AT), very often occur close together in time. The correlation coeffi- cients between them are 0.98–0.99 on aver- age, and air temperature from 12 UTC is Figure 1. Polish cities analyzed in this study about 1–1.3°C lower than the maximum air temperature. 56 Magdalena Kuchcik

Table 3. Values of the 95th and 90th percentiles of Apparent Temperature (AT) (b), percentile values at Tmax=30°C (c), AT values corresponding to percentiles in column c (d) and the yearly mean air temperature (e) in 10 Polish cities, 1993–2002

ab cde

AT percentile [°C] AT value Mean air

Tmax=30°C at Tmax percentile temperature T th th Cities 95 90 Tmax percentile (column c) [°C] Białystok 26.2 23.6 0.985 29.8 7.4

Gdańsk 23.2 20.6 0.996 30.3 8.6

Kraków/Cracow 26.9 24.2 0.980 29.8 8.6

Lublin 26.4 24.0 0.987 30.7 7.8

Poznań 26.5 23.8 0.979 28.7 8.9

Rzeszów 27.7 24.8 0.978 30.3 8.4

Szczecin 26.2 23.3 0.982 29.6 9.1

Toruń 26.7 23.9 0.977 29.3 8.6

Warszawa/ 26.8 24.2 0.981 30.2 8.4 Warsaw Wrocław 26.8 24.2 0.978 29.1 9.2

Source: Author’s own elaboration

The application of the ‘Apparent Tem- resulted in only 13 days with heat waves in perature’ definition only reduces slightly all analyzed cities over a 10 year. These in- the number of heat waves in the analyzed cluded 9 days in Szczecin in which two heat cities (69 as opposed to 74 with the air tem- waves were generated in the same year, one perature threshold 30°C definition). How- of 4 days (July 23–26, 1994) and another ever these lasted 9 days on average, more of 3 (July 30 to August 1, 1994) (Table 4). than twice as long as the heat waves defined Moreover, Szczecin distinguishes itself from using the threshold of 30°C (4 days). The other Polish cities by the highest dew tem- ‘hottest’ heat waves according to this defi- perature, relatively high humidity and a low nition occurred in Toruń, a city situated in wind speed. the center of Poland on the Vistula River Most heat waves that are defined in line and characterized by a greater frequency with the 3-day threshold of 30°C definition of hot days than the surrounding area. In last exactly 3 days and are usually included line with the ‘30°C’ definition, the ‘hottest’ in the ‘Apparent Temperature’ defined heat waves are noted in Poznań. However, ones, sometimes with two being included Toruń has vet-higher values for Apparent in one of the second type. For example, in Temperature. While the second definition Warsaw, only 3 of 11 defined heat waves is far more restrictive and resulted in ‘hot- lasted longer than 3 days, in Lublin—just ter’ yet shorter heat waves, it is interesting one (Table 5). In a few cases the dates of that the minimum air temperature is almost the heat waves defined in these two differ- equal in the 2 types. At the Polish seaside ent ways do not even overlap. The applica- there were no heat waves defined as at least tion of Humidex didn’t result. Does it mean 3 days with air temperature above 30°C. that there are no longer periods perceived Applying the Humidex threshold of 40°C by people as heat? It means rather that the Defining Heat Waves… 57

Table 4. Heat wave statistics for Poland over the years 1993–2002 in relation to three definitions.

Humidex Heat waves Min. 6 days with AT>95th percentile Min. 3 days with T > 30°C max >40°C

No.of Mean No. of Mean No. of heat dura- heat dura- No. of heat

Cities waves tion Tmax Tmin AT waves tion Tmax Tmin AT days waves

Białystok 7 9.9 29.4 15.0 29.4 6 4.8 31.9 14.8 31.0 - -

Gdańsk 4 13.8 24.4 18.4 25.7 - - -

Kraków/ 8 7.6 29.9 15.8 28.4 10 4.0 32.2 16.2 30.4 - - Cracow Lublin 5 10.0 29.2 16.2 29.7 6 4.2 31.6 16.3 32.3 2 -

Poznań 6 8.3 30.2 15.4 29.3 6 4.5 33.0 16.2 30.7 1 -

Rzeszów 7 8.6 29.7 14.4 30.1 10 4.5 31.7 14.7 32.7 - -

Szczecin 10 9.8 28.6 14.7 29.1 8 4.6 32.2 15.7 32.0 9 2

Toruń 7 8.3 30.4 15.9 30.6 9 4.2 32.8 17.1 32.4 - -

Warszawa/ 8 7.9 30.0 16.7 30.3 11 4.0 31.8 16.5 31.6 - - Warsaw Wrocław 7 8.6 30.0 14.2 29.4 8 4.9 32.2 14.7 30.2 1 -

Mean 6.9 9.3 29.2 15.7 29.2 8.2 4.4 32.2 15.8 31.5 - -

AT – Apparent Temperature, T – air temperature, Tmax, Tmin – maximum and minimum air temperature Source: Author’s own elaboration

threshold of air temperature could have types defining them, would seem to be by been inappropriate for the entire area of analyzing heat waves in relation to mortal- Poland because obviously at many loca- ity and/or morbidity. Such analyses have al- tions the level of 30°C is exceeded on just ready been undertaken by the author (Kuch- a few days each year. cik 2001, Kuchcik in Kozłowska-Szczęsna et al. 2004, Kuchcik and Blażejczyk 2005). The issue of heat waves is especially valid CONCLUSION when account is taken to IPCC projections pointing to an increased frequency or magni- This survey of the approaches to heat tude of extreme events such as heat waves in waves presented in this paper make it clear a warmer global environment (WMO 2001). just how important to particular studies the An example of such a scenario was brought to proper choice of the definition is. Generally, Poland in summer 2006, in which there was a at the basis of defining heat waves lies the heat wave lasting almost one month, 2 months idea that several consecutive very hot days of drought and a flood at the end. However, are more burdensome to people than sepa- even where precisely-calculated indices for rate hot days. The effect of the ‘wave’ should mortality during past periods are at our dis- thus be tested through detailed weather– posal, one major unknown to be taken account health studies. This is why the best way to of as future impact: is estimated is whether or specify heat waves in a given location and not the population will able to adapt and accli- to indicate adequate thresholds or air-mass matize to changes in climate (WHO 2003)? 58 Magdalena Kuchcik

Table 5. Dates and air temperature characteristics of heat waves in Poland, 1993—2002, as defined in two different ways.

Min. 6 days with AT>95th percentile Min. 3 days with Tmax> 30°C

No. No.

of Tmax Tmin AT of Tmax Tmin AT City Start End days [°C] [°C] [°C] Start End days [°C] [°C] [°C]

25 July 94 7 Aug .94 14 32.5 14.1 30.8 12 July 94 16 July 94 5 30.7 12.9 29.0

27 May 95 2 Juue 95 7 27.7 12.2 27.3 25 July 94 2 Aug. 94 9 33.5 13.5 31.6

6 June 98 11 June 98 6 28.6 14.7 29.7 4 Aug. 94 6 Aug. 94 3 32.2 14.3 31.1

Białystok 14July 01 30 July 01 17 27.8 16.4 28.9 21 June 00 23 June 00 3 31.4 14.1 29.3

16 Aug. 01 22 Aug.01 7 29.6 14.7 30.5 15 July 01 17 July 01 3 31.8 18.7 34.9

10 July 02 19July 02 10 29.0 17.0 29.5 28 Juyl 02 2 Aug. 02 6 31.7 15.6 29.9

28 July 02 4 Aug. 02 8 30.6 16.2 29.4

21July 94 10 Aug.94 21 25.6 18.9 27.3

Gdańsk 17 Aug. 97 3 Sep. 97 18 25.1 16.6 24.9 _ 31 July 02 6 Aug. 02 7 23.8 20.0 25.9

14 Aug. 02 22 Aug.02 9 23.2 18.1 24.6

24 July 94 8 Aug. 94 16 32.0 15.3 29.2 3 Aug. 93 5 Aug. 93 3 31.4 14.2 29.6

9 July 95 16 Juyl 95 8 29.0 16.2 28.4 26 June 94 29 June 94 4 31.1 15.5 29.8

18 Aug. 95 24 Aug.95 7 29.1 13.6 27.7 25 July 94 3 Aug. 94 10 33.1 15.0 29.9

Kraków/ 7June 96 12 June 96 6 29.5 14.7 27.8 5 Aug. 94 7 Aug. 94 3 32.0 16.0 29.5

Cracow 16 Aug. 00 21 Aug.00 6 31.6 17.1 30.3 7 June 96 10 June 96 4 30.4 14.1 28.6

15 Aug. 01 20 July 01 6 29.7 16.7 28.7 28 June 97 30 June 97 3 32.7 17.3 31.3

18 June 02 23 June 02 6 29.7 15.7 27.8 20 July 98 23 July 98 4 32.6 17.0 30.9

28July 02 2 Aug. 02 6 28.7 17.5 27.4 4July 99 6 July 99 3 32.1 16.3 30.3

19 Aug. 00 21 Aug.00 3 33.5 18.8 32.4

14 July 01 16 July 01 3 33.1 18.4 31.3

24 July 94 8 Aug. 94 16 31.8 15.6 30.3 25 July 94 3 Aug. 94 10 32.8 15.9 31.1

5 June 98 11 June 98 7 27.7 15.5 29.4 5 Aug. 94 7Aug. 94 3 31.7 15.2 30.3

Lublin 15 Aug. 01 22 Aug.01 8 28.5 15.7 29.2 8 June 96 10 June 96 3 30.3 14.4 30.6

7 July 02 19 Jjuly 02 13 28.1 17.2 29.2 21 July 98 23 July 98 3 31.5 17.3 34.4

28 July 02 2 Aug. 02 6 29.8 17.3 30.5 19 Aug. 00 21 Aug.00 3 31.9 16.8 33.0

15 July 01 17 Julyl 01 3 31.6 18.0 34.3

21 July 94 6 aug 94 17 32.4 16.0 30.5 24 July 94 2 Aug. 94 10 34.3 17.0 32.6

8 July 95 16 jul 95 9 29.7 16.6 29.3 8 July 95 10 July 95 3 31.7 15.3 30.5

7 June 96 12 jun 96 6 30.2 15.3 29.5 7 June 96 9 June 96 3 31.9 14.8 30.5

Poznań 10 Aug. 97 15 aug 97 6 28.2 13.4 27.9 20 June 00 22 June 00 3 35.0 15.0 31.9 Defining Heat Waves… 59

Poznań 22 Aug. 97 27 aug 97 6 28.9 13.7 28.0 9 July 02 11 July 02 3 32.9 17.8 27.6

28 July 02 2 aug 02 6 31.7 17.5 30.8 28 July 02 1 Aug. 02 5 32.5 17.2 31.4

23 July 94 9 Aug. 94 18 31.9 13.5 31.2 27 June 94 30 June 94 4 31.5 14.6 32.4

9 July 95 15 July 95 7 28.8 13.9 29.8 24 July 94 8 Aug. 94 16 32.3 13.6 31.5

Rzeszów 7 June 96 12 June 96 6 30.2 10.7 29.6 8 June 96 10 June 96 3 31.6 11.0 31.4

6 June 98 11 June 98 6 29.1 15.5 30.8 20 July 98 23 July 98 4 31.8 15.3 33.6

15 Aug. 01 21 Aug.01 7 29.4 13.9 30.7 4 July 99 6 July 99 3 31.2 14.0 32.8

7 July 02 18 July 02 12 29.4 16.7 29.4 21 June 00 23 June 00 3 31.7 14.3 32.5

28 July 02 2 Aug. 02 6 29.1 16.3 29.5 19 Aug. 00 21 Aug.00 3 33.2 16.9 34.8

14 July 01 16 July 01 3 32.1 15.9 34.2

16 Aug. 01 18 Aug.01 3 30.1 14.5 31.6

14 July 02 16 July 02 3 31.4 16.6 32.1

21 July 94 6 Aug. 94 17 33.4 17.1 34.8 11 July 94 15jJuly 94 5 32.1 14.8 31.2

7 July 95 16 July 95 10 29.2 15.8 30.3 21 July 94 2 Aug. 94 13 34.3 17.8 36.3

27Jjuly 95 7 Aug. 95 12 28.4 13.4 28.5 8 July 95 10 July 94 3 31.0 15.1 32.2

Szczecin 18 Aug. 95 24 Aug.95 7 29.9 13.5 29.5 29 July 95 31 July 95 3 30.6 15.2 31.3

7 June 96 12 June 96 6 28.5 14.8 30.3 18 Aug. 95 20 Aug.95 3 30.5 13.0 30.7

19 aug 96 24 aug 96 6 28.3 14.6 27.9 20 June 00 22 June 00 3 33.2 14.5 32.1

8 Aug. 97 15 Aug.97 8 27.3 14.4 27.4 9 July 02 11 July 02 3 32.7 17.1 28.7

22 July 01 30 July 01 8 26.6 13.9 27.0 29 July 02 1 Aug. 02 4 32.7 18.1 33.1

19 Aug. 01 26 Aug.01 8 26.9 15.9 27.9

16 Aug. 02 31 Aug.02 16 27.7 14.1 27.3

23 July 94 6 Aug. 94 15 33.8 16.3 32.4 12 July 94 15 July 94 4 32.2 16.2 30.8

7 July 95 15 July 95 9 30.4 15.3 30.4 24 July 94 3 Au.g 94 11 34.6 16.5 33.2

17 Aug. 95 24 Aug.95 8 29.8 13.5 31.0 8 July 95 10 July 95 3 33.0 15.7 32.6

Toruń 7 June 96 12 June 96 6 30.0 15.4 29.9 7 June 96 9June 96 3 31.7 15.6 31.6

25 July 01 30 July 01 6 27.6 15.4 27.6 28 June 97 30June 97 3 33.1 19.0 32.6

15 Aug. 01 22 Aug.01 8 29.9 18.1 31.3 4 July 99 6 July 99 3 32.1 18.5 32.7

28 July 02 2 Aug. 02 6 31.4 17.0 31.5 20 June 00 22 June 00 3 35.0 15.9 32.2

15 Aug. 01 17 Aug.01 3 32.0 18.4 33.3

29 July 02 2 Aug. 02 5 31.8 18.4 32.4

11 July 94 17 July 94 7 29.7 16.8 29.8 13 July 94 16 July 94 4 30.7 17.3 30.7

23 July 94 7 Aug. 94 16 32.6 16.6 32.1 24 July 94 2 Aug. 94 10 34.0 16.3 33.1

8 July 95 14 July 95 7 29.0 15.3 29.6 4Aug. 94 6 Aug. 94 3 31.7 16.9 32.5

Warszawa/ 18 Aug. 95 24 Aug.95 7 30.0 14.1 29.6 19 Aug. 95 21 Aug.95 3 30.5 14.6 30.4

Warsaw 6June 98 11June 98 6 29.2 17.2 29.3 8 June 96 10 June 96 3 31.4 15.4 30.4

15 Aug. 01 22 Aug.01 8 29.4 16.9 31.0 28 June 97 30June 97 3 31.5 17.9 31.8 60 Magdalena Kuchcik

Min. 6 days with AT>95th percentile Min. 3 days with Tmax> 30°C

No. No.

of Tmax Tmin AT of Tmax Tmin AT City Start End days [°C] [°C] [°C] Start End days [°C] [°C] [°C]

Warsaw 13 July 02 18 July 02 6 28.8 18.6 29.8 4 July 99 6 July 99 3 31.8 16.4 31.8

28 July 02 2 Aug. 02 6 31.6 18.1 31.2 20 June 00 22 June 00 3 32.9 14.2 30.7

14 July 01 16 July 01 3 32.8 17.7 33.4

19 Aug. 01 21 Aug.01 3 30.3 17.1 31.3

28 July 02 2 Aug. 02 6 31.6 18.1 31.2

22 July 94 7 Aug. 94 17 33.2 15.3 31.1 26 June 94 29 June 94 4 32.2 16.1 32.3

8 July 95 16 July 95 9 29.7 16.1 29.3 13 July 94 15July 94 3 31.2 14.3 29.6

18 Aug. 95 24 Aug.95 7 29.4 13.9 28.8 22 July 94 6 Aug. 94 16 33.5 15.2 31.3

Wrocław 7 June 96 12 June 96 6 30.5 14.4 30.1 7 June 96 9 June 96 3 31.4 13.1 30.7

22 Aug. 97 27 Aug.97 6 28.9 12.4 28.3 20 July 98 22 July 98 3 32.2 15.7 30.3

14 Aug. 00 21Aug. 00 8 30.0 14.1 29.7 20 June 00 22 June 00 3 33.4 13.7 31.3

25 Aug. 02 31Aug. 02 7 28.4 13.3 28.7 9 July 02 11 July 02 3 32.3 15.6 27.2

28 July 02 31 July 02 4 31.6 14.1 28.9

Source: Author’s own elaboration

REFERENCES treme Chicago heat waves, Climate Research, 24:243–254 AMS (American Meteorological Society), Glos- Davis, R., Knappenberger, P., Michaels, P. and sary of Meteorology, http://amsglossary.allen- Novicoff, W. (2003), Changing Heat-Related press.com/glossary/search?p=1&query=hea Mortality in the United States, Environmental t+wave Health Perspectives, 111, 14: 1712–1718. Baranowska, M., Czarniecki, W., Grzędziński, De’Donato, F., Michelozzi, P., Kalkstein, E. (1968), Badania wpływu bioklimatycznie L., D’Ovidio, M., Kirchmayer, U., Accetta, G. uciążliwych dni lata (z uwzględnieniem dni and Perucci, C.A. (2005), The Italian Project parnych, gorących i upalnych) na dolegli- for Prevention of Heat-Health Effects During wości podmiotowe osób zdrowych i chorych Summer, Findings from 2005, 17th Interna- z nadciśnieniem tętniczym [The Study on the tional Congress of Biometeorology, Annalen Influence of Bioclimatologically Burdensome der Meteorologie, 41, 1: 287–290. Summer Days (sultry, hot and very hot days) European Climate Assessment & Dataset project on the Ailments of Healthy and with Hyper- ECA&D, http://eca.knmi.nl/indicesextremes. tension Persons], Polskie Archiwum Medycyny Ellis, F.P. and Nelson, F. (1978), Mortality in the Wewnętrznej, 41, 3/9: 317–324. Elderly in a Heat Wave in New York City, Beniston, M. and Stephenson, D.B. (2004), Ex- August 1975, Environmental Research, 15: treme Climatic Events and their Evolution 504–512. under Changing Climatic Conditions, Global Environment Canada, Heat Wave Duration Index, and Planetary Change, 44: 1–9. http://www.hazards.ca/maps/trends/wsdi- Changnon, D., Sandstrom, M., Schaffer, Ch. e.html. (2003), Relating changes in agricultural Fortuniak, K., Kłysik, K. and Wibig, J. (2006), practices to increasing dew points in ex- Urban-rural contrasts of meteorological Defining Heat Waves… 61

parameters in Łódź, Theoretical and Applied Kalkstein, L.S.and Valimont, K. (1986), An Eval- Climatology, 84: 91–101. uation of Summer Discomfort in the United Johnson, H., Kovats, R.S., McGregor, G., Sted- States using a Relative Climatological Index, man, J., Gibbs, M. and Walton, H. (2005), Bulletin of the American Meteorological Soci- The Impact of the 2003 Heat Wave on Daily ety, 67, 7: 842–848. Mortality in England and Wales and the Use Katsouyanni K., Pantazopoulou A., Toiloumi G., of Rapid Weekly Mortality Estimates, Euro- Tselepidaki I., Moustris K., Asimakopou- surveillance, 10, 7–9: 168–171. los D., Poulopoulou G. and Trichopoulos D. Huynen, M.M.T.E., Martens, P., Schram, D., Wei- (1993), Evidence for Interaction between Air jenberg, M.P. and Kunst, A.E. (2001), The Pollution and High Temperature in the Cau- Impact of Heat Waves and Cold Spells on sation of Excess Mortality, Archives of Envi- Mortality in the Dutch Population, Environ- ronmental Health 48, 235–242. mental Health Perspectives, 109, 5: 463–470. Keatinge, W.R., Donaldson, G. C., Cordioli, E., IPCC (Intergovernmental Panel on Climate Martinell, M., Kunst, A., Mackenbach, J., Change) (2001), Synthesis Report, Glossary of Nayha, S. and Vuori I. (2000), Heat Related Terms, http://www.grida.no/climate/ipcc_tar/ Mortality in Warm and Cold Regions of Eu- vol4/english/204.htm. rope: Observational Study, British Medical Kalkstein, L.S. (1991), A New Approach to Eval- Journal, 321: 670–673. uate the Impact of Climate on Human Mor- Koppe, Ch. and Jendritzky, G. (2004), Final tality, Environmental Health Perspectives, 96: Report of the Meeting of Experts to Develop 145–150. Guidelines on Heat-Health Warning Systems, Kalkstein, L.S. (2004), Heat and Health: Meth- Freiburg, Germany, annex VI, pp. 4. odological Considerations for Warning Sys- Koppe, Ch., Kovats, S. and Jendritzky, G. (2003), tem Development, oral presentation on the Heat Health Warning Systems, unpublished workshop “Per non morire di caldo”—Climate workshop draft, Workshop on Vulnerability to Changes, Heat Waves and Population Health, Thermal Stresses, Freiburg, Germany, pp. 25. Organized by the Municipality of Rome, Kosatsky, T. (2005), The 2003 European Heat March 30, 2004. Waves, Archives : Eurosurveillance monthly Kalkstein, L. S., Barthel, C. D., Greene, J. S. and releases, 10, 7–8: 148–149. Nichols M. C. (1996a), A New Spatial Synop- Kozłowska-Szczęsna, T., Krawczyk, B. and Kuch- tic Classification: Application to Air Mass cik, M. (2004), Wpływ środowiska atmosfe- Analysis, International Journal of Climatol- rycznego na zdrowie i samopoczucie czło- ogy, 16: 983–1004. wieka, [The Influence of the Atmospheric Kalkstein, L.S. and Corrigan, P. (1986), A Synop- Environment on Human Health and Wellbe- tic Climatological Approach for Geographi- ing], Monografie, IGiPZ PAN, 4, Warszawa, cal Analysis: Assessment of Sulfur Dioxide pp. 194. Concentrations, Annals of the Association of Kuchcik, M. (2001), Mortality in Warsaw: Is American Geographers, 76, 3: 381–395. There Any Connection with Weather and Air Kalkstein, L.S. and Greene, J. (1997), An Evalu- Pollution? Geographia Polonica, 74, 1: 29–45. ation of Climate/Mortality Relationships in Kuchcik, M. (2006), Fale upałów w Polsce Large U.S. Cities and the Possible Impacts of w latach 1993–2002 [Heat Waves in Poland a Climate Change, Environmental Health Per- over the years 1993–2002], Przegląd Geogra- spectives, 105, 1: 84–93. ficzny, 78, 3: 397–412. Kalkstein, L.S., Jamason, P.F., Greene, J.S., Libby, Kuchcik, M. and Błażejczyk K. (2005), Regional J. and Robinson, L. (1996b), The Philadelphia Differentiation of Heat Waves in Poland and Hot Weather-Health Watch/Warning System: their Impact on Mortality. DWD, Annalen Development and Application, Summer 1995, der Meteorologie, 41, 1, 415–418. Bulletin of the American Meteorological Soci- Kyselý, J. (2004), Mortality and Displaced Mortal- ety, 77, 7: 1519–1528. ity during Heat Waves in the Czech Republic, 62 Magdalena Kuchcik

International Journal of Biotemeteorology, 49, Rooney, C., McMichael, A., Kovats, S. and Cole- 2: 91–105. man, M. (1998), Excess Mortality in England Kyselý, J. and Huth, R. (2004), Heat-Related and Wales, and in Greater London, During Mortality in the Czech Republic Examined the 1995 Heat Wave, Journal of Epidemiology through Synoptic and ‘Traditional’ Approach- and Community Health, 52: 482–486. es, Climate Research, 25: 265–274. Sartor, F., Snacken, R., Demuth, C. and Walcki- Larsen, J. (2003), Record Heat Wave in Europe ers, D. (1995), Temperature, Ambient Ozone Takes 35 000 Lives, Earth Policy Institute, Oc- Levels, and Mortality During Summer, 1994, tober 9, 2003–9, www.earth-policy.org/Up- in Belgium. Environmental Research, 70: dates/Update29_printable.htm 105–113. Lascheski, G. and Jendritzky, G. (2002), Effects Semenza, J., Rubin, C., Falter, K., Selanikio, J., of the Thermal Environment on Human Flanders, W. D., Howe, H. and Wilhelm J. Health: An Investigation of 30 Years of Daily (1996), Heat-Related Deaths during the July Mortality Data from SW Germany, Climate 1995 Heat Wave in Chicago, The New Eng- Research, 21: 91–103. land Journal of Medicine, 335, 2: 84–90. Matzarakis, A. and Mayer, H. (1991), The Ex- Sheridan, S.C. (2002), The Redevelopment of treme Heat Wave in Athens in July 1987 from a Weather-Type Classification Scheme for the Point of View of Human Biometeorology, North America, International Journal of Cli- Atmospheric Environment, 25B, 2: 203–211. matology, 22: 51–68. McGeehin, M.A. and Mirabelli, D. (2001) The Smoyer, K. (1998), A Comparative Analysis of Potential Impacts of Climate Variability and Heat Waves and Associated Mortality in St. Change on Temperature-Related Morbidity Louis, Missouri: 1980 and 1995, International and Mortality in the United States. Environ- Journal of Biometeorology, 42, 1: 44–50. mental Health Perspectives, 109 (Supplement Smoyer, K., Rainham, D. and Hewko J. (2000) 2): 185–189. Heat-Stress-Related Mortality in Five Cities Michelozzi, P., De Donato, F., Accetta, G., Fo- in Southern Ontario: 1980–1996, Internation- rastiere, F., D’Ovidio, M., Perucci, C., Kalk- al Journal of Biometeorology, 44: 190–197. stein, L. S. (2004), Impact of Heat Waves on Smoyer-Tomic, K. and Rainham, D. (2001), Beat- Mortality—Rome, Italy June–August 2003, ing the Heat: Development and Evaluation of Morbidity and Mortality Weekly Report, 53, 17: a Canadian Hot Weather Health-Response 369–371. Plan, Environmental Health Perspectives, 109, Michelozzi, P., Fano, V., Forastiere, F., Barca, 12: 1241–48. A., Kalkstein, L.S. and Perucci, C. (2000), Souch, C. and Grimmond, C.S.B. (2004), Applied Weather Conditions and Elderly Mortality in climatology: ‘heat waves’, Progress in Physical Rome during Summer, World Meteorological Geography, 28, 4: 599–606. Organization Bulletin 49, 4: 348–355. Szymanowski, M. (2005), Interactions between Niedźwiedź, T. (ed.) (2003), Słownik Meteorolo- Thermal Advection in Frontal Zones and the giczny [Meteorological Dictionary], Instytut UrbanHeat Island of Wrocław, Poland, Theo- Meteorologii i Gospodarki Wodnej, Polskie retical and Applied Climatology, 82: 207–224. Towarzystwo Geofizyczne, Warszawa. Steadman, R.G. (1979), The Assessment of Sultri- Oke, T.R. (1987), Boundary Layer Climates, Lon- ness. Part I: A Temperature-Humidity Index don, Methuen, pp. 435. Based on Human Physiology and Clothing Pattenden, S., Nikiforov, B. and Armstrong, B.G. Science. Part II: Effects of Wind, Extra Ra- (2003), Mortality and Temperature in Sofia diation and Barometric Pressure on Apparent and London, Journal of Epidemiology and Temperature, Journal of Applied Meteorology, Community Health, 57: 628–633. 18: 861–885. Robinson, P. (2001), On the Definition of a Heat Steadman, R.G. (1984), An Universal Scale of Wave, Journal of Applied Meteorology, 40, 4: Apparent Temperature, Journal of Applied 762–775. Meteorology, 23, 12: 1674–1687. Defining Heat Waves… 63

UNEP (United Nations Environment Programme) WHO (2003), The Health Impacts of the 2003 (2004), Impacts of Summer 2003 Heat Wave in Summer Heatwaves, Briefing Note for the Europe, www.grid.unep.ch/product/publica- Delegations of the 53 Session of the WHO tion/download/ew_heat_wave.en.pdf. Regional Committee for Europe, pp.12 http:// Thompson, D.R., Pohl, J.E., Yiu-Yu, S. T., Hiorns, www.euro.who.int/document/Gch/HEAT- R.W (1996), Meteorological Factors and the WAVES%20RC3.pdf Time of Onset of Chest Pain in Acute Myo- WMO (2001), Final report of the Meeting of Ex- cardial Infarction, International Journal of perts to Develop Guidelines on Heat-Health Biometeorology, 39, 2: 116–120. Warning Systems—www.wmo.ch/web/wcp/ Thornbrugh, C. (2001), Are America’s Cities clips2001/html/HHWS_docs/Heat-Health% Ready for the Hot Times Ahead? Heat Wave 20Warning%20Systems.htm). Awareness Project, www.isse.ucar.edu/heat/ casestudy.html Paper first received: June 2006 Watts, J. and Kalkstein, L. (2004), The Develop- In final form: July 2007 ment of a Warm-Weather Relative Stress In- dex for Environmental Applications, Journal of Applied Meteorology, 43: 503–513.

CHANGES OF THE VISTULA RIVER CHANNEL AND FLOODPLAIN IN THE LAST 200 YEARS

ADAM ŁAJCZAK*, JOANNA PLIT**, ROMAN SOJA***, LESZEK STARKEL*** AND JUSTYNA WAROWNA**** *Faculty of Earth Sciences, University of Silesia, Będzińska 60, 41-200 Sosnowiec, Poland E-mail: [email protected] **Institute of Geography and Spatial Organization, Polish Academy of Sciences, Twarda 51/55, 00-818 Warsaw, Poland E-mail: [email protected] ***Institute of Geography and Spatial Organization, Polish Academy of Sciences, Sw Jana 22, 31-018 Kraków, Poland E-mail: [email protected]; E-mail: [email protected] ****Institute of the Earth Sciences, M.Curie-Skłodowska University, Al. Kraśnicka 2cd, 20-718 Lublin E-mail:[email protected]

Abstract: The Vistula River is a typical Central-European river flowing from the mountains across basins and upland belts to the lowlands. The Vistula valley is modelled by a river with a complex hydrological regime. In its upper reaches, floods driven by summer rainfall prevail, while in the lower reaches snowmelt floods are important. Deforestation favoured a natural propensity for river braiding. In the mid-19th century, the channelization of the upper Vistula (in the Car- pathian foreland) and the lower reaches was commenced with, while the middle streach was left in a natural state, such that the river has in places preserved a braided pattern up to the present day. The channelization followed by construction of reservoirs caused downcutting and aggra- dation to occur, such that opposing tendencies were observed in particular reaches of the river channel. In addition, flood embankments confined aggradation to the intra-embankment area. Thus, the functioning of the Vistula River system is largely controlled by diverse human activity. Unconstrained flow and river load transport along the whole river length are only partly possible during extreme floods. The present-day adjustment tendencies also relate to ongoing changes in land-use in the drainage basin, as well as on global climatic changes.

Key words: Vistula River, channelization/regulation, present-day changes of floodplain, downcut- ting, aggradation. 66 Adam Łajczak, Joanna Plit, Roman Soja, Leszek Starkel and Justyna Warowna

INTRODUCTION omorphologists, of which one is interested in studying the impact of regulation works on The Vistula River is one of the great central geomorphic processes (J.Warowna) and the European rivers, flowing from the moun- other focuses on present-day sediment loads tains in the South (the Alps, Sudetes, Car- (A.Łajczak). pathians) towards the North or Baltic Seas. In contrast to other European rivers All these rivers traverse a series of morpho- (Petts et al. 1989), the Vistula had a specific logical units, starting from the mountain feature in the last few centuries. Its catch- headwaters and submontane basins, cross- ment area once constituting the central part ing an upland belt and, then, entering an of the Polish state was divided, in the late extensive lowland zone which, in the east- 18th century, between the three neighbour- ern part, was several times occupied by the ing countries of Austria, Russia and Prussia. Scandinavian ice sheet. In the catchments Only in 1918 did it come back under Polish of the Elbe, Oder and Vistula rivers, we rule. Water management in the three coun- find the remains of former streamways, later tries in question differed greatly. Prussia was crossed by rivers flowing straight to the sea. the first to start intensive regulation works Another characteristic feature of this zone is in the lower course of the Vistula in the a gradual change of hydrological regime mid-19th century. At about the same time, from one dominated by rainy floods in the after several major floods in the Upper Vis- West to one with more pronounced snowmelt tula catchment, the Austrian government floods towards the East. Therefore, in their started channelization works there. The upper reaches especially, the river floods are longest (middle) reach of the Vistula River, driven by heavy summer rainfall. left under Russian rule, remained practically The evolution of the Vistula River valley unchanged (except in local fragments) until in the last 10–15,000 years has been studied independence was regained. Even now, this in detail under IGCP Project 158, which segment of the river preserves many fea- focused on the palaeohydrology of the tem- tures inherited from previous centuries (e.g. perate zone (Starkel et al. 1991), as well as braided channels and islands). This triparti- on the role of human impact in the last few tion is still visible in present-day processes millennia. It was the largest catchment and and forms, as being superimposed upon the the longest river studied within this Project. natural sequence of processes along the lon- The results were published in six volumes gitudinal profile of the Vistula. (Starkel, ed. 1982–1996). This paper tends to continue those stud- ies but concentrates on the processes of and CHARACTERISTICS OF THE VALLEY, trends to the evolution of the Vistula river FLOODPLAIN AND CHANNEL channel and floodplain. Its aim is to char- acterize these during the last 100–200 years The Vistula River drainage basin, with an of direct human impact, by focusing on the area of 199 813 km2, is drained by a 1047 km metachronous regulation of the river course long river. In what is an asymmetric drainage and the transformation of water discharge basin, right-bank tributaries predominate in and the sediment load. It therefore supple- both the upper reaches (Carpathian tribu- ments previous studies. Specifically, this pa- taries) and the middle reaches (Figure 1). per results from multi-specialist cooperation The Vistula flows across numerous of geographers, namely a geomorphologist morphotectonic units of diversified relief concentrating on the role of humans in long- (Table 1). Particular sections of its valley dif- term trends (L. Starkel), a hydrologist work- fer in depth and width, the extent of the flood- ing on the Vistula River regime (R.Soja), plain, channel width and gradient (Starkel a cartographer-analyst of river-channel 1990). A short Carpathian section of a nar- changes over time (J. Plit), and two other ge- row valley cut in flysch rocks has a gradient Changes of the Vistula River Channel and Floodplain… 67

Figure 1. The Vistula River catchment against a background of geomorphic regions 1. main watershed; 2. water reservoirs; 3. hydrological stations mentioned in this paper (1. Skoczów, 2. Goczałkowice, 3. Smolice, 4. Tyniec, 5. Sierosławice, 6. Jagodniki, 7. Szczucin, 8. Sandomierz, 9. Zawichost, 10. Annopol, 11. Puławy, 12. Warszawa, 13. Kępa Polska, 14. Włocławek, 15. Toruń, 16. Tczew). a—Carpathians; b—Subcarpathian basins, c—Polish Uplands, d—Polish Lowlands (Mazovian part), e—Polish Lowlands (Kuyavian-Pomeranian part) 68 Adam Łajczak, Joanna Plit, Roman Soja, Leszek Starkel and Justyna Warowna : Starkel (ed.)(1990) Table 1. Morphometric characteristics of Vistula Valley reaches Valley 1. Morphometric characteristics of Vistula Table Source Changes of the Vistula River Channel and Floodplain… 69 of 14.6‰ on average. As the Vistula turns carpments widens and narrows. The straight eastward, (in the Oświęcim Basin), the val- or winding river channel is only 300–500m ley broadens, and the winding or meandering wide, yet is regulated and embanked. The channel has a width varying from 20 to 50m. river flow is controlled by Włocławek Res- In the section considered, several tributaries ervoir. The 45 km long outlet section of the merge with the Vistula. Over the next 35km regulated Vistula crosses the reclaimed area section, the river flows across limestone of the delta. The marked variability in relief horsts of the Cracow Gate, then along the of the Vistula valley—as controlled by lithol- northern margin of the Sandomierz Basin ogy and tectonics, as well as the deglaciation where it is shifted by the Carpathian tributar- pattern and spatial pattern of larger tributar- ies. Here, in the river valley (8 to 25km wide), ies—has an unquestionable influence on the the Holocene alluvial plain is 3 to 10km wide. processes which have modelled and model its A regulated and embanked winding Vis- floor (Starkel 1990). tula channel (once a meandering one), with a width of 100–200m tends to occupy the middle of the valley floor, albeit undermin- HYDROLOGICAL CHARACTERISTICS ing the river left-bank plateaus in certain sec- OF THE VISTULA tions. At the confluence point with the river San, the Vistula turns northward and enters The complex hydrological regime of the Vis- a gap across the Polish Uplands cut into tula is related to the location of its drainage limestones and gaizes. In the narrowed sec- basin in the transient temperate climatic tions of the valley, the 1–2km wide floodplain zone. The dominance of a continental or occupies the whole floor, while in broader maritime influences varies across an annual basin-shaped sections it is up to 10 km wide. cycle, resulting in marked year-to-year vari- The up to 1km wide channel has been braid- ability in discharges. Depending on precipi- ed before the last few decades. The Vistula tation distribution in the drainage basins of then turns north-west and flows towards the the tributaries, hydrological parameters of Warsaw Basin (Mazovian Lowland), where the Vistula increase first rapidly but in an it is joined by the river Narew and then turns irregular manner to Zawichost (gauging sta- westward. Here, the 0.5–1km wide braided tion no. 9, see Figure 1), and then gradually Vistula channel uses one of the two mar- in the downstream reach of the river. ginal depressions. Downstream of the Bzu- Following the contribution by Fal et al. ra river, (a left-bank tributary), the Vistula (1997), the hydrological characteristics are starts its lower Kuyavian-Pomeranian reach based on data from four gauging stations incised in deposits from the last glaciation. located along the Vistula River (Figure 2, The river channel gradient declines gradu- Table 2). In winter, the river discharges in- ally to 0.1‰ and the valley confined by es- crease steadily in the whole Vistula drainage

Table 2. Characteristic flows [m3/s] at four hydrological stations along Vistula River, 1951–1990

Gauging station Qav Qav Nov.-Apr. Qav May-Oct Qmax1% Qmin1% Nowy Bieruń 20.7 20.9 20.6 804 1.29 Sandomierz 292.0 294.0 289.0 7,500 54.70

Warszawa 573.0 614.0 533.0 7,400 102.00

Tczew 1,080.0 1,230.0 932.0 9,190 243.00

Source: Fal et al. 1997. 70 Adam Łajczak, Joanna Plit, Roman Soja, Leszek Starkel and Justyna Warowna

Figure 2. Mean monthly discharges of the Vistula River at four hydrological stations Source: after Fal et al. 1997

basin from November through to the culmina- sandy covers store rainwater, and thus in the tion of meltwater flow. From May, discharges downstream section winter-season outflow decrease along the Vistula’s whole length, yet predominates over summer. the flow is always greater than the annual av- The amplitude in Vistula water levels erage. In the summer months (June–August), results from the presence of flood embank- the second peak flow is related to extreme ments, which cause the water level to rise summer floods in the Carpathians. It progress- in the intra-embankment area. In the Car- es along the Vistula to Sandomierz (gauging pathian reach of the Vistula, without em- station no.8, see Figure 1) and then ceases bankments and with a steep gradient, the in the middle and lower reaches of the river. water level amplitude is 2m, and cf. 3.5m in The lowest flows occur in September and are the foothill reach. On the other hand, in the related to precipitation deficit and the small area of Subcarpathian Basins this amplitude storage capacity of the drainage basin. Sum- varies from 5 to 8.5m, depending on the gap mer–autumn droughts take in the whole Vis- between the embankments. Downstream of tula drainage basin. The mean annual flows the San river mouth, in the broad channel, increase from 6m3/s in Skoczów to 21m3/s in the amplitude of water level diminishes to Nowy Bieruń, 292m3/s in Sandomierz,573m3/s 6–7m in Warsaw, while in Tczew it reaches in Warszawa and 1080m3/s in Tczew (Figure its highest value of 10m. Over the whole 2). Winter flows predominate by 1% over length of the Vistula, the highest recorded summer-season flows to Nowy Bieruń, far- water levels coincide with the heights of the ther downstream by 7%, and finally by 14% in embankment. Under natural conditions, (i.e. Tczew. This pattern reflects variable sources prior to straightening and deepening of the of water supply and a diminishing influence Vistula channel), water-level amplitude was of the Carpathians downstream. In the Car- possibly only half as great (Makowski and pathian part of the drainage basin the river Tomczak 2002). regime is controlled by heavy summer rain The Vistula is a river with high and fre- storms, with accompanying overland flow. quent summer floods triggered by continu- Rainfall diminishes northward, permeable ous rainfall in the Carpathians and their Changes of the Vistula River Channel and Floodplain… 71 foreland (Figure 3). The flood of 1813, which ever, the channelization of the Vistula River occurred across the whole Vistula drain- in the last 150 years, as well as thermal and age basin, is believed to have been the most chemical pollution of the water, reduced ice- spectacular in the whole period for which jamming. On the other hand, frazil-ice-jams observations were carried out. The floods of began to form in backwater areas of the res- 1844, 1903, 1839, and 1934 were of a slightly ervoirs. smaller magnitude. The flood of 1997 was The highest observed Vistula discharg- 3 characterized by the highest recorded dis- es reached 7450m /s at Zawichost (Q1% is charges at the majority of the Vistula’s wa- 8160m3/s at Zawichost, according to Punzet ter-level gauging stations to Zawichost. The 1991) and 8000m3/s at Tczew. However, if the analysis of water levels in the lower reach at first peak flow resulted from a rainfall flood, Toruń (gauging station no.15, see Figure 1) the second was due to a snowmelt flood. since 1770 points to the occurrence of some The 1% water flow corresponds to a spe- years without significant floods (5–8 cases in cific runoff of 460 l/s/km2 at Nowy Bieruń, a century), as well as a few years recording 236 l/s/km2 at Sandomierz, only 88 l/s/km2 2 or 3 high floods (Makowski and Tomczak at Warsaw and 47 l/s/km2 at Tczew. The fre- 2002). Spring floods, which occur regular- quency of floods (p%) was calculated using ly, predominate, while summer and winter the log Pearson type III distribution applied floods are less frequent. by Polish hydrologists (Punzet 1991). In case Flood waves formed in the Carpathian of the Vistula basin the extreme floods in the part of the Vistula drainage basin reach upper and middle course of the river are gen- their maxima at Sandomierz or Zawichost, erated by summer rainy floods in the Car- before being quickly flattened downstream pathians and downstream the Narew junc- such that only occasionally does the peak tion mainly by snowmelt floods. flow increase to the river mouth (as in 1962, In the 20th century, the mean annual Figure 3). The summer flood of 1997 had outflow from the Vistula drainage basin did a discharge of 2000m3/s in Cracow, 5830m3/ not show statistically significant trends (Fal s downstream of the Dunajec river mouth, 1993). Studies on changes in the flood re- 5800m3/s in Sandomierz, 4730m3/s in War- gime of the Vistula River and its major tribu- saw, and 4220m3/s at Tczew (Barczyk et al. taries have been performed many times. The 1999). The increase in maximum discharge obtained negative regression coefficients re- values, resulting from superimposed flood ferred to peak flow and the frequency and waves of the Vistula and its tributaries, usu- duration of floods in the years 1901–1970 ally occurs in the Sandomierz Basin. The (Stachy and Nowak 1977). Similar results peak discharges might be altered through were found when examining the variability embankment breaking and the formation of to extreme annual discharges in the years an episodic flood lake. An inundated area 1921–1992 (Stachy et al. 1996). In the medi- in excess of 300 km2 was formed during the um-size catchments of the Carpathian part summer flood of 1934. of the Vistula basin, there is a trend towards Extreme snowmelt floods (Figure 4) oc- a decrease in flood magnitude (Soja 2002), cur less frequently than the rainfall-induced but there is a steady increase in minimum ones, though in this case the maximum dis- discharges related to reservoir management. charges increase steadily down the river. In general, the hydrological regime of the In 1979, the peak discharges of the Vistula Vistula River in the 20th century has been were 1290m3/s at Sandomierz, 2550m3/s at characterized by a decrease in the number Warsaw and as much as 7020m3/s at Tczew, of floods. Nevertheless, catastrophic floods the result of a superimposed flood wave of have occurred, some of these being clus- the Vistula and the thawing flood waves in tered in subsequent years, but then followed the Bug and Narew drainage basins. Ice-jam by 15–20-year periods without major floods floods occurred very often in the past. How- (Starkel 2001). 72 Adam Łajczak, Joanna Plit, Roman Soja, Leszek Starkel and Justyna Warowna

Figure 3. Maximum discharges in the longitudinal profile of Vistula River during summer rainy floods Source: based on Soja and Mrozek in Starkel (ed.) 1990

Figure 4. Maximum discharges in the longitudinal profile of Vistula River during snowmelt floods. Source: based on Soja and Mrozek in Starkel (ed.) 1990 Changes of the Vistula River Channel and Floodplain… 73

PROCESSES IN THE REGULATED VISTULA the river channel (Figures 5, 6). In the years VALLEY—CHANGES IN THE PATTERN AND 1800–1980, the upper Vistula was shortened SHAPE OF THE RIVER CHANNEL by 34,3km in the Oświęcim Basin (Czaja et al.1993) and by 35,5km in the Sandomierz THE UPPER REACH: FROM SPRINGS Basin (Trafas 1992). TO THE SANDOMIERZ BASIN The channelization works started in vari- Apart from in the short mountainous sec- ous reaches of the Vistula at different times. tion, the Vistula valley become deforested, In the Carpathian part only one reservoir farmed and densely populated. Analysis (Czarne), with a capacity of 4.5 million m3, of the three maps dating from the turn of was built in 1974. A series of concrete sills the 18th century to the mid-19th century1, have been constructed to prevent downcut- shows that the Vistula formed its channel ting. The next reservoir (Goczałkowice), on freely on a valley floor of diverse width. At the flat floor of the Oświęcim Basin, is up that time, it was a meandering, unregulated to 14m deep and 11km long, and accumu- river (excluding the section within the city lates all bedload and a majority of suspend- of Cracow). The river upstream of the Cra- ed load. In the area of the Oświęcim Basin cow Gate had meanders of small (50–200m) the Vistula preserved its winding character, radii, the channel being 140–170m wide.2 despite the channelization in the 1920s, Downstream of Cracow to the Dunajec out- when the river channel was shortened by let, the Vistula channel was very winding, 40%. As a result the channel gradient in- with small-radii meanders (300–400m). creased from 0.34‰ to 0.56‰ (Czaja et al. The old maps do not depict larger bars and 1993) and downcutting has been activated. natural levees. Downstream of the Dunajec In the 1950s, the construction of a cascade river mouth, the Vistula was 400m wide, of dams with canal locks in the region of bars occurred in the channel and the mean- Cracow inhibited the process of intensive der radii were larger (500–600 m). downcutting which has been observed since From the beginning of 19th century, the the mid-19th century. Ingarden (1922) es- section upstream of Sandomierz was char- timated the channel deepening at 2.5m, acterized by accumulation, the channel be- while Punzet (1981) reports the value of 3m came shallower and wider, and numerous for the period of 1871–1954. Such extensive bars and islets were formed. In the mid-19th erosion was supported by the cutting and century, the channel changed gradually in filling of the abandoned channels (these the reach between the mouths of the Duna- playing the role of bypass channels), and by jec and San rivers. Although the Vistula exploitation of sand from the river bed up still flowed in a sinuous channel, the river to 1950. was shallower; it occupied a wider chan- In the Sandomierz Basin, channelization nel and had left numerous bars. Numer- works entailing cutting of river bends started ous distributaries and abandoned channels in 1848. At the same time, the channel was occurred. The crucial transformation was narrowed by a system of dykes and groynes. related to embanking and straightening of The gradient thereby increased from 0.28 to 0.32‰ (Trafas 1992). In the second half 1 Detailed Map of the Polish Kingdom prov- of the 20th century, the erosion rate of the inces of 1783 (Mapa szczegulna wojewodztw ko- Vistula channel bottom was ca. 1 cm/year. ronnych) of Karol de Perthès, scale 1: 225 000, The channel sinuosity in the Upper Vis- Regna Galiciae et Lodomeriae Josephi II et M. tula dropped from 1.7 to 1.4 (Babiński and Theresiae... of 1790 of Joseph Liesganig, scale Klimek 1990). The downcutting is reported 1: 28 800 and Carte von West-Gallizien.. of 1808 of Mayer von Heldensfeld, scale 1:172 000. to reach from 0.5 to more than 1m (Ingar- 2 Measurements based on Übersichts Plane den 1922, Punzet 1981). The cutting works Weischel Stromes… 1:14 400, manuscript map of resulted in river bends of large curvature— 1851–1852. extending the radius to 1km (Trafas 1992). Figure 5. Times of regulation of the Vistula River channel Source: elaborated by J.Plit and J.Warowna Figure 6. Times of embankment of the Vistula River Source: elaborated by J.Plit 76 Adam Łajczak, Joanna Plit, Roman Soja, Leszek Starkel and Justyna Warowna

The substantial narrowing of the flood chan- limited to short, non-channelized reaches nel resulted in repetitive breaking of flood (locally reaching 20–100m in 1980–2003) embankments and the formation of crevasse and to the formation of bank failures, which splays (Gębica et al. 1998). are several meters deep and tens of meters long, to the leeward of dykes. Sandy mate- THE VISTULA IN THE GAP THROUGH rial transported as bedload was accumulated THE UPLANDS between the artificial structures in the intra- The Vistula valley in the gap section (see groynes areas. Based on the volume calcu- Figure 1 and Table 1) was continually de- lation, the average sedimentation rate was forested from the Middle Ages on. Channel 1.3 million m3 per year. The polders have mobility depended on valley width. At sites filled up to 2m above mean water level and where the valley was narrower, the river was a new man-made terrace is being formed. confined in the same location. Still at the This floodplain terrace may be up to 350m beginning of the 19th century, the Vistula wide. Its flat surface is slightly inclined to- meandered in certain sections, and flowed wards the river or raised as a levee-like in two or three parallel winding branches. form near the channel, and, then, descends The main channel shifted from one bank towards the natural flood plain where back- to the other. Numerous islets stabilized by swamps occur. The intra-groyne sedimenta- vegetation were formed at that time. The tion is accompanied by downcutting in the abandoned meanders or flood channels of active channel axis. Evaluation of the inci- the Vistula were often used by the tributar- sion rate is hindered by migration of the ies (Kalicki and Plit 2003). bed bars, which causes fluctuations in the Changes in river character occurred in floor heights of up to 2m per year (Warowna stages, especially after the series of large 2003). The presence of groynes results in the floods in the 19th century. The river depos- formation of potholes up to 12m deep. ited numerous bars, formed islands, and oc- cupied an extensive area. The width of the THE VISTULA VALLEY meandering channel was 100–200m, while IN THE MAZOVIAN REACH that of the braided channel was 300 –1500m After leaving the gap, the Vistula used to trans- (Warowna 2003). The average annual rate form its main channel, chiefly in the locations of bank erosion was 3–9m. where the valley widened. As recently as in the In the 19th century, an increase in the 18th century, the Vistula was still meandering size of particles deposited on the floodplain in the central section up to the confluence with ((from mud to sand) was recorded (Falkowski the Narew. After the flood of 1813, the Vistula 1967), and a change from a sinuous chan- used 2, or more rarely 3, branches. In certain nel to a wide, shallow and straightened one sections it became a braided river with chan- with shifting bars. This transformation re- nel width reaching 1000m (Plit 2004). Over sulted from accelerated outflow of water the last 250 years, the width of the belt where and sediment load through the channelized the channel shifted was 1–4 km (Falkowski reaches of the Upper Vistula and tributar- 1967). This process was limited artificially ies (Warowna 2003). The sedimentation of to the intra-embankment width in the 19th sand in the upper part of the gap reached 1m century (Figure 6). The narrowed sections of (Falkowski 1967). The channelization works the valley are related to the presence of rocks performed partly in the 1920s and later be- resistant to erosion (clays mainly) and are tween 1960 and 1990 resulted in a narrowing characterized by a deeper channel and strong of the mean water level channel (by a system lateral erosion. The embanking of the river of dams and groynes) to 120–300m and to was performed successively, section by sec- the cutting off of side branches by dykes tion. It started after the catastrophic floods in (Figure 7). The width of the bankfull chan- 1844 and 1888/1889, but the undertaking was nel is 500–600m. The lateral erosion was ended after the Second World War. Changes of the Vistula River Channel and Floodplain… 77

Figure 7. Changes in channel and floodplain form caused by the regulation in the middle course of the Vistula River between 1960 (A) and 1995 (B). 1—embankments, 2—various regulation constructions, 3—bars revegetated before 1960, 4—active channel bars, 5—bars revegetated after new regulation works. Source: elaboreted by J. Warowna

Antierosional protection was provided by longitudinal works and rip-raps, so the for bridge pillars, industrial plants and the channel narrowed from 500–700m to 340m. urbanized area of Warsaw. In this reach, the Successively, particular side branches were lateral erosion may rise to 4–5m per year. cut off, so the whole flow concentrated in In Warsaw channelization started after the one main channel. Due to highly resist- flood of 1884, when the river channel shift- ant clay deposits exposed in the bottom, ed 500m eastward. The reach was regulated downcutting does not occur despite the fact 78 Adam Łajczak, Joanna Plit, Roman Soja, Leszek Starkel and Justyna Warowna that the channel is narrowed (Jacewicz and against the ongoing erosion. Accumulation Kuźniar 2000). forms appear just 25km down from the dam. In the 18th and 19th centuries, the lower The material building those forms is much Vistula downstream of the Narew mouth was coarser when compared with that known a braided river with bars and islands. Two from the pre-reservoir years. or rarely three parallel troughs functioned The formation of an ice cover, flow of ice in short sections in the 19th century. On floes and ice jams are independent factors the valley floor the series of troughs, which modelling the bottom and banks of the chan- evidence a parallel shifting of the river, are nel (Grześ and Banach 1983) The shores preserved. Abandoned meanders occur only of Włocławek Reservoir are deformed by locally, in places where the valley widens. In thermal expansion of ice causing ice-plough this area, which was under Russian rule up ridges to form. Those ridges can be several to 1918, the first flood control structures and tens of centimetres high and up to 1.5m regulation works had been initiated by the wide (Banach 1994). The ice floes drifting end of the 19th century (Babiński 1992). downstream damage the unregulated banks, The reach downstream of Warsaw, where while ice-jams divert the current towards singular groynes do not modify the braided the distributary troughs, reactivating them pattern of the river, was not channelized or triggering downcutting downstream of a until the 1960s. At present, the mean flow jam. It is believed that still-existing unregu- channel has been narrowed to 300m. The lated river reaches are ice-jam generating. primary width of the braided river bed was 1.1–1.6km, with a mean water table width THE VISTULA VALLEY of 0.6–0.9km. The patterns of the side bars IN THE KUYAVIAN-POMERANIAN REACH and river current vary and 60% of the bank The first regulation of the partly braided lengths are eroded. Migration of the chan- and partly anastomozing lower Vistula in nel bars follows at a rate of 0.6–1.4m/day the areas formerly under Prussian rule was (Babiński 1992). However, the construc- performed in the 18th century. The hydro- tion of the reservoir in Włocławek in 1969 technical works comprised the construction modified the ongoing processes. The back- of groynes and dykes or rip-raps, the clear- water affects a 58km section upstream of ing of the willow thickets from the channel, Włocławek. The amplitudes in the water the levelling of the bottom and narrowing of table reach 2.2m. The lateral erosion has the channel, the shifting of the river bed and give way to abrasion and mass movements the construction of a canal connecting the (affecting 45% of bank lengths), which were Vistula and Oder rivers. Modernization and most intensive in the first years just after banking up ended in 1880–1892. reservoir construction. The edge of the Up to the end of the 19th century, it was morainic plateau retreated by 150–200m accumulation processes that occurred, then in places. After the ice-jam flood of 1982, regulation followed. Channel deepening was the low inundated islands have been re- intensified after the shortening of the outlet moved and up to 1.5m high sandy-gravel to the Baltic Sea in 1840, and in the years protective benches have been set up along 1895–1915. The deforested floor of the Vis- 15% of the banks (Banach 1994). The reach tula valley was then used as farmland and below Włocławek was channelized in the meadows. Frequent catastrophic floods in 1950s, but ca. 25% of the banks are eroded the 19th century caused people to abandon at present (Banach 1998). Downstream of settlement and cultivation in the floodplain. the reservoir, downcutting reached 2–3m The Lower Vistula was channelized sys- in the first 20 years, over a distance of 3km tematically in the last 25 years of the 19th from the dam. Downstream erosion is ceas- century (Figure 4). The regulatory measures ing gradually (Babiński 1992). Armouring were miscellaneous. The channel was nar- of the bottom with cobbles is a protection rowed to 350 –375m, but in the delta both Changes of the Vistula River Channel and Floodplain… 79 branches were narrowed to 250m and 125m, flow of Vistula water. Higher embankments respectively. Accumulation of the sandy-silty were built and the intra-embankment area deposits in polders between the groynes and broadened. The original natural outflow on the regulation structures reach 3.4– 4.2m routes were abandoned (Makowski 1997). above the river level (Szmańda 2000). The Once development work ended, river alluvia channel capacity decreases, because the ceased to fertilize the delta. Accumulation deposition at the channel banks is larger than in the form of an underwater delta occurs at the volume of the eroded material. The man- the bottom of the Gulf of Gdańsk. made terrace is 375–450m wide (Babiński 1992), and is higher in the narrowed parts of THE ROLE OF FLOOD EMBANKMENTS the channel. The formation of levees 2–20m Construction of the flood embankments wide and up to 1m high, which separate de- and confining of the peak flood flow zone cantation pools up to 1.8m deep and 36m has been carried out progressively since the wide, is observed. The mean depths of the mid-19th century (Figure 6), although long channel upstream of Fordon were 1.6–2.0m embankments have been present in Żuławy (at widths of 730–780m) prior to channeliza- since the beginning of the 14th century (Ma- tion, but the minimum depths decreased to kowski 1997). In the Upper Vistula, the gap 0.7m. Since the mid-1970s, depths have been between the embankments is of 600–800m reaching 3.0m. As a result of the channeliza- (Czajka 2000), increasing to 900–1500m tion, this section of the river is straight, with downstream of the junction with the San. alternating bars and pools. However, this zone shrinks to 600m occa- sionally, due to bridging, or to 450m, if the THE VISTULA DELTA old embankments are located closer to the The Vistula delta has been deforested river channel. In the Mazovian Lowland, for centuries, subjected to intensive farming, the width of the intra-embankment area and dissected by a dense network of artificial is 1000–1700m, excluding within Warsaw drainage canals. Both the rivers and the ca- where it is narrowed to 400–600m (Jace- nals were embanked. At the beginning of the wicz and Kuźniar 2000). Downstream of 19th century, the Vistula entered the Baltic Włocławek, the gap between the embank- sea using three branches. The majority of ments is 1000–1500m. Reduction of the flow was concentrated in the right branch. flood sedimentation area to 25–40% of the According to Majewski (1969), in the 19th primary one causes fast aggradation in the century, this branch of the river used to shift intra-embankment area, and the formation the landline 25–30m offshore per year. The of a new terrace level in the channelization embankments confined accumulation to the zone (Warowna 2003). In the Oświęcim intra-embankment area and to the fans, ex- Basin, accretion of a 1.5m thick layer has plaining why the Vistula channel is located been noted since the 1950s (Czajka 2000). higher than the surrounding plains. The del- If the embankments are broken, as during ta was inundated very often. Snowmelt floods July 1997, flood crevasse troughs up to 300m were particularly hazardous, as the flood long and 10m deep form, as well as crevasse wave entered the still-frozen lower Vistula splays to one kilometre long and 0.7m thick and the Gulf of Gdańsk. In February 1840, (Gębica et al. 1998). the left branch of the river was ice-jammed. The flood wave formed a new outlet route ca.14km shorter, which started to form a fan- SEDIMENT TRANSPORT, DEPOSITION delta quickly, while the accretion of the old AND LOADS delta was hindered. Some hazardous floods (especially ice-jammed ones) in the 1880s, The north-facing slope of the Carpathians is forced a 9km long shortcut and the forma- the major source area of material supplied tion of an artificial outlet allowing 90% out- to the river as dissolved and suspended loads Figure 8. Differentiation of mean annual suspended load R (in tons) along the Vistula River between 1946 and 1995. Changes of the Vistula River Channel and Floodplain… 81 and as bedload (Brański 1974; Maruszczak ervoirs. The contribution of suspended load 1984; Łajczak 1999). The role of the Polish to the overall transport of clastic material Uplands in this supply is less significant, the shows a downward trend downstream in the lowland area making a minor contribution. Upper Vistula, and varies from 90% to 70%. The average suspended load transport in the In the lowland tributaries, suspended-load Carpathian, upland and lowland parts of the transport amounts to 50%, if 23–28% in the Vistula drainage basin is of 200, 10 and 3t/ lower Vistula, according to Babiński (1992). km2 per year respectively, while the average The examination of the grain-size composi- dissolved loads are 100, 60 and 45t/km2 per tion of deposits in the Carpathian reservoirs year (Maruszczak 1984; Łajczak 1999). Only confirms the dominance of clay-silty parti- in the mountain area of the Vistula drain- cles (Łajczak 1999). In Włoclawek Reser- age basin does suspended load and bedload voir, the silty-clay deposits only dominate in transport predominate. In upland and low- the deepest parts. land terrain, the level of ionic transport is The mean annual suspended-load trans- several times greater than suspended-load portation in the Vistula increases suddenly transport. In the mountain and upland trib- at the mouth of tributaries, and then de- utaries of the Vistula and the stretch of the creases between river junctions (Figure 8). main river in the Carpathian foreland does Among the Carpathian tributaries, the San suspended material dominate the bedload supplies the largest load to the Vistula, i.e. load (Brański and Skibiński 1968; Froehlich 0.8 million tons per year. The greatest trans- 1982). port of suspended load occurs in the stretch The rate of suspended-load transport of- of the Vistula downstream of the San river fers only indirect evidence of erosion inten- mouth, where the average load was ca 1.6 sity in a catchment. Downstream in a river, million tons in the second half of the 20th any estimation of erosion magnitude in the century. The magnitude of the transported drainage basin becomes more and more dif- load prior to dam construction was esti- ficult. The losses in transported suspended mated at 2.3 million tons (Łajczak 1999). load are expressed by reference to the sedi- In the lower reach of the Vistula, prior to ment delivery ratio (DR), which indicates construction of the Włocławek Dam (in the percentage of entrained material which 1968), the magnitudes of transported loads flows out at a gauging site. This ratio de- were steadily graded over the whole length creases downstream, and in sub-catchments of the river. After the reservoir came into of the order of 102–104 km2 in area it reaches use, the magnitude of the load decreased to from a few to a dozen percent (Maruszczak 0.5 million tons. 1984). In the mountain catchments this The magnitude, seasonal and multi-an- ratio is higher. In the small catchment of nual pattern of transport and sedimenta- a stream, cart tracks and forest ways can sup- tion in the Vistula have changed drastically ply about 80% of the suspended load to that due to human impact. Significant changes stream (Froehlich 1982; Froehlich and Wall- occurred in the 20th century due to chan- ing 1992). On the scale of the whole drain- nelization works and the setting up of hydro- age basin of the Vistula, DR does not exceed technical structures (Łajczak 1995, 1999). 1% as a result of the presence of many res- The total supply from the tributaries was

a—hypothetical sediment load (without depositional effects), b—real sediment load S—mean annual deposition (in tons/yr and tons/ha/yr. in the balanced reaches (calculation based on difference between successive measuring points). On the longitudinal axis indicated 16 hydrological stations (see Figure 1) and position of the dam in Włocławek. Source: elaborated by A.Łajczak 82 Adam Łajczak, Joanna Plit, Roman Soja, Leszek Starkel and Justyna Warowna of ca 3.6 million tons per year in the second The suspended load settles partially on half of the 20th century, the upland tributar- the floodplain during floods (in the intra- ies supplying 0.4 million tons per year and embankment area at present) and in reser- the lowland ones 0.5 million tons. voirs. The average annual losses in trans- In the annual cycle, suspended-load ported load attest to intensified overbank transport in the Upper Vistula predomi- sedimentation and have been calculated for nates in the summer months (June–August) successive river reaches between the gauging over the spring months (March–April). stations, according to the input-output meth- The influence of the mountain tributar- od, for the period 1946–1995. They show an ies decreases downstream. Therefore, in upward trend along the whole Upper Vistula the middle reach of the Vistula, the early (Figure 8). These losses diminish consider- spring and summer transported load are ably downstream, and only beyond the Na- similar, while in the lower reach it is spring rew mouth and in the backwater stretches loads that predominate. Downstream of of the Włocławek Dam are they larger. the dams, transported loads have started to Accepting that, since the beginning of the even out during a whole year. 20th century it is channel deepening, initi- The effects of human impact are most ated by channelization, that predominates, noticeable when multi-annual changes in the accumulation of suspended load can suspended load transport are analyzed. Up- only be related to the intra-embankment stream of the Cracow surroundings large area. The overbank accumulation increases changes in transport in the second half of along the Upper Vistula and reaches almost the 20th century were related to the supply 1000 tons/ha beyond the San confluence. of physical pollutants to the river (mainly This fact is confirmed by the complete filling from coal mines via the Przemsza river). of abandoned channels, which were cut off The size of the suspended load increased there at the beginning of the 19th century, up to 1970, reaching double the value and by accretion of the Vistula floodplain in of the period 1946–1950 (from ca 0.2 to the gap stretch (Maruszczak 1982; Łajczak 0.4 million tons respectively). Up to the 1999). In contrast, the morphology of older mid-1980s, suspended-load transport de- abandoned channels of the Vistula located creased rapidly, the following years bringing upstream is still visible. In effect, shoaling of a stabilization at a lower level than before this reach of the Vistula takes place and re- 1950 (Łajczak 1999). Downstream of Cra- sults in prolonged overbank flooding leading cow, up to the mouth of the river, the sus- to greater floodplain sedimentation (Łajczak pended load transport showed a downward 1999; Warowna 2003). In further stretches trend. This is related to a larger number of of the Vistula, to its mouth, overbank sedi- reservoirs set up and to changes in land-use mentation decreases to ca 10 tonnes per ha (a reduction in the area of arable fields). per year, and just between the Narew river The most significant decrease, by a fac- outlet and the dam in Włocławek it exceeds tor of six, occurred in the lower reaches of 50 tons per ha per year. Repeated surveying the San river in 1946–1995, as an effect of indicates that the accumulation of suspend- significant changes in land use within the ed load in the intra-embankment area occurs river catchment post 1947. Despite human at the highest rate in the zones occupied by induced changes, the influence of hydrocli- natural levees, and at the lowest rate in the matic factors on the multi-annual pattern vicinity of the flood embankments. to suspended load outflow is still evident. Two reservoirs on the Vistula River play In the Carpathian tributaries and along the an important role in sediment storage. The whole course of the Vistula to its mouth, al- Goczałkowice Reservoir, stopped storing most decade-long fluctuations in suspended suspended load durably after 20 years in load transport result from the major floods operation, though it captures the entire bed- which occur every few years. load. The Włocławek Reservoir stores half Changes of the Vistula River Channel and Floodplain… 83 of the incoming suspended load and the en- eral increase in transport and in overbank tire bedload. The influence of this reservoir sedimentation due to channel deepening has on the transport of suspended load in the been observed. In the following decades, the lower reach of the Vistula has varied over rates of transportation and sedimentation time, depending on reservoir shoaling and declined as more deep reservoirs were con- dragging. Small dams on the Vistula River structed. The channelization and simultane- near Cracow hardly stop suspended load at ous intensified chemical pollution caused all. The most significant reduction in sedi- the natural Vistula River to take the role of mentation, even by a factor of five, has been a canal directing the water surplus out of the noted downstream of the mouths of tributar- drainage basin. The diversity of channeliza- ies on which deep reservoirs are located. tion works undertaken in the second half of the 20th century, combined with the setting up of reservoirs and dams on the Vistula and CONCLUSIONS its tributaries, brought about changes in the size of transported and deposited material, From its mountain reach, via the fore- and in effect, aggradation in the intra-em- mountain basins and the upland belt, as well bankment zone at many locations, the forma- as through the lowland zone down to the tion of a man-made terrace and an exceed- delta at the outlet to the Baltic Sea, the Vis- ance of maximum water stages hazardous to tula valley is currently modelled by a river the whole floodplain (Figure 7). The role of whose hydrological regime is complex. Be- major floods has become more important, sides the summer floods related to extreme so the unconstrained flow and river load rainfall occurring in the mountains, thawing transport over the whole river length are floods also occur and are often combined possible only during extreme floods, while with ice-jams (such floods becoming more aggradation can occur on the whole flood- significant downstream in the river). The plain. The milder winter seasons (with a floods triggered in the mountains lose their lesser hazard due to thawing floods) seem to impetus in the foreland, where the inundat- influence the transportation of the river load ing rivers leave behind a considerable load of positively. Ice-jam floods occur rather rarely, transported material. due to severe water pollution and a rise in The significant deforestation of the water temperature. On the other hand, the drainage basin prior to the 19th century led power-generation function of the reservoirs to a change from a meandering to a braid- has a negative influence on flood control ed channel pattern in the Upper and Mid- safety. The Vistula floods downstream of dle Vistula and to a change from an almost the Włocławek Reservoir which occurred in anastomosing or winding pattern to a braid- 1982 and at the beginning of April 2006 are ed one in the lower reaches. Ice-jam floods, good examples here. especially in the middle reach, have resulted The changes in land-use (especially re- in channel shifting and avulsion in widened duction in arable area) observed in the last sections. 20 years favour decreases in suspended-load Channelization started in the mid-19th transport and in aggradation. The summer century in the areas which were formerly floods favour channel-deepening in the up- under Austrian rule (the upper reach) and per reaches of the rivers, but simultaneously Prussian rule (the lower reach). This result- increase the hazard of sudden floods down- ed in channel braiding in the middle reach, stream. This may lead to changes in valley- as well as progressing delta aggradation and floor management, and to more rigorous simultaneous channel deepening in regulat- compliance with water management rules. ed reaches (Figure 9). Since the beginning The widening of the intra-embankment of more extensive channelization in the first zone and withdrawal of settlements and in- three decades of the 20th century, a gen- frastructure from flood areas will become Figure 9. Channel regulation, embankments and erosional-depositional tendencies in the longitudinal profile of Vistula River during last two centuries. 1—water reservoirs (time of construction), 2—water steps and weirs, 3—main periods of channel regulation and construction of dam, 4—new outlets of Vistula River to the Baltic Sea, 5–9—main trends of fluvial processes in three time transects: before flood series in 1840s at the beginning and at the end of 20th century; 5—distinct tendency to downcutting, 6—downcutting prevailing over aggradation, 7—aggradation prevailing over downcutting, 8—distinct tendency to aggradation, 9—river section transformed to canal. Source: elaborated by L. Starkel Changes of the Vistula River Channel and Floodplain… 85 necessary, as will reconstruction of natural Barczyk, A., Bogdanowicz, E., Dobrzyńska, I., channel patterns and riverside plant com- Drezińska, B., Fal, B., Kruszewski, A., Ośród- munities. Extensive polders for temporary ka, K. and Szturc, J. (1999), Przebieg wezbra- flood water storage will need to be built, es- nia w dorzeczu Wisły (Course of Flood in the pecially if we consider the increase in flood Vistula River Catchment Area), in Grela, J., frequency during the progressive climate Słota, H. and Zieliński, J. (eds.), Dorzecze Wi- change. All these tasks are foreseen, but may sły. Monografia powodzi lipiec 1997, Instytut not reach the investment stage. Restoration Meteorologii I Gospodarki Wodnej (IMGW), of fluvial ecosystems through reduction of 45–76. the input of sewage and industrial pollutants Brański, J. (1974), Ocena denudacji dorzecza has to progress simultaneously. Wisły na podstawie wyników pomiarów ru- When comparing the Vistula with those mowiska unoszonego (Evaluation of Denuda- Western or central European rivers whose tion Process in the Vistula Catchment Area/ headwaters are located in mountains (exam- Drainage Basin in the Light of Measurements ples might be the Rhine, Elbe, Oder, or Dan- of the Suspended Loads), Prace IMGW, Insty- ube (Petts et al. 1989), we can conclude that tut Meteorologii i Gospodarki Wodnej, 6. Vistula never attained any uniform system Brański, J. and Skibiński, J. (1968), Udział rumo- of channel regulation. The river still bears wiska wleczonego w ogólnej masie rumowiska the 19-century traces of the different models rzecznego transportowanego w korycie środ- of water management applied independently kowej i dolnej Wisły (Participation of Bed by each of the three countries then occupy- Load in General Mass of Stream Bed Load ing partitioned Polish territory. For each of Transported in the Middle and Lower Vistula these countries, the Vistula played only the Channels), Wiadomości Służby Hydrologicz- a marginal role of a frontier river. nej i Meteorologicznej, 3–4. Czaja, S., Degórska, V. and Leśniok, M. (1993), Naturalne i antropogeniczne zmiany biegu REFERENCES koryta Wisły od zbiornika w Goczałkowicach do ujścia Przemszy (Natural and Man-Made Babiński, Z.(1992), Współczesne procesy kory- Changes in the Channel Course of theVis- towe dolnej Wisły (The Present-Day Fluvial tula River form the Goczałkowice Reservoir Processes of the Lower Vistula River), Prace to the Mouth of the Przemsza river), Geo- Geograficzne, Instytut Geografii i Przestrzen- graphia, Studia et Dissertationes, Uniwersytet nego Zagospodarowania (IGiPZ), PAN 157, Śląski, 17, 7–15. 171pp. Czajka, A. (2000), Sedymentacja pozakorytowa Babiński, Z. and Klimek, K.(1990), The Present- aluwiów w strefie międzywala Wisły w Kotli- Day Channel and Flood-Plain of the Vistula nie Oświęcimskiej (Overbank Sedimentation River, Geographical Studies (Special issue: in the Inter-Embankment Zone of the Vistula Evolution of the Vistula River Valley During River in the Oświęcim Basin), Przegląd Geo- the Last 15 000 years, Part III,), 5: 62–75. logiczny, 48, 3, 263–267. Banach, M.(1994), Morfodynamika strefy brze- Fal, B. (1993), Zmienność odpływu z obszaru gowej zbiornika Włocławek (Morphodynam- Polski w bieżącym stuleciu (Variability in ics of the Włocławek Reservoir Shore Zone), Outflow from Poland’s Territory in the 20th Prace Geograficzne, Instytut Geografii i Century), Wiadomości IMGW, Instytut Mete- Przestrzennego Zagospodarowania (IGiPZ), orologii i Gospodarki Wodnej, 16(37), 3: PAN, 61, 180pp. 3–20. Banach, M.(1998), Dynamika brzegów dolnej Fal, B., Bogdanowicz,E., Czernuszenko,W., Wisły (Dynamics of the Lower Vistula River Dobrzyńska, I. and Koczyńska, A. (1997), Banks), Dokumentacja Geograficzna, Insty- Przepływy charakterystyczne głównych rzek tut Geografiii Przestrzennego Zagospodaro- polskich w latach 1951–1990 (Characteristic wania (IGiPZ), PAN, 9: 9–39. Flows in Main Poland’s Rivers in the years 86 Adam Łajczak, Joanna Plit, Roman Soja, Leszek Starkel and Justyna Warowna

1951–1990), Materiały Badawcze, Seria: Hy- grafii i Przestrzennego Zagospodarowania drologia i Oceanologia, 21, 174pp. PAN 19:17–30. Falkowski, E. (1967), Ewolucja holoceńskiej Wi- Kalicki, T. and Plit, J. (2003), Historical Changes sły na odcinku Zawichost-Solec i inżyniersko- of the Vistula Channel and its Reflection in -geologiczna prognoza jej dalszego rozwoju the Floodplain Between Józefow and Kazi- (Holocene Evolution of the Vistula River mierz Dolny. Holocene and Late Vistulian Valley between Zawichost and Solec and an Paleogeography and Paleohydrology, Prace Engineering Geological Prognosis of its Fu- Geograficzne 189: 179–207. ture Development), Biuletyn PIG, Państwowy Klimek, K. (1983), Erozja wgłębna dopływów Wi- Instytut Geologiczmy, 198: 57–142. sły na przedpolu Karpat (Erosion In-Depth of Froehlich, W. (1982), Mechanizm transportu the Vistula Tributaries on the Fore-Carpathi- fluwialnego i dostawy zwietrzelin do koryta w an Area), in Kajak, Z. (ed.), Ekologiczne pod- górskiej zlewni fliszowej (The Mechanism of stawy zagospodarowania Wisły i jej dorzecza, Fluvial Transport and Waste Supply into the Państwowe Wydawnictwo Naukowe (PWN), Stream Channel in the Mountainous Flysch Warszawa, 97–108. Carpathians), Prace Geograficzne, Instytut Łajczak, A.(1995), The Impact of River Regula- Ggeografii i Przestrzennego Zagospodaro- tion, 1850–1990, on the Channel and Flood- wania, PAN,143. plain of the Upper Vistula River, Southern Froehlich, W. and Walling, D. (1992), The Use of Poland, in Hickin, E.J. (ed.) River Geomor- Radionuclides in Investigation of Erosion and phology, Wiley & Sons, Chichester, 209–233. Sediment Deliver in the Polish Carpathians, Łajczak, A.(1996), Modelling the Long-Term in Erosion, Debris Flows and Enverinment Course of Non-Flushed Reservoir Sedimen- in Mountain Regions, Proceedings of the tataion and Estimating the Life of Dams, Chengdu Symposium, July 1992, IAHS Pub- Earth Surface Processes and Landforms, 21. lications, 209: 61–76. Łajczak, A.(1999), Współczesny transport i se- Gębica, P., Patkowski, B., Lasek, A. and Sokołow- dymentacja materiału unoszonego w Wiśle ski, T. (1998), Geomorfologiczne i sedymen- i głównych dopływach (Contemporary Trans- tologiczne skutki przerwania wałów przeciw- port and Sedimentation of the Suspended powodziowych w dolinie Wisły w lipcu 1997 Material in the Vistula River and its Main roku (Geomorphic and Sedimentological Ef- Tributaries), Monografie, Komisja Gospodar- fects of a Break in the Embankments in the ki Wodnej PAN, 15, Warszawa, 215pp. Vistula valley in July 1997), in Materiały z Łajczak, A.(2003), Contemporary Transport of konferencji ‘Powódź w dorzeczu górnej Wisły w Suspended Material and its Deposition in the lipcu 1997 roku’ , Kraków, 185–194. Vistula River, Poland, in Hydrobiologia, Klu- Grześ, M. and Banach, M. (1983), Powódź zatoro- wer Academic Publishers, Netherlands. wa na Wiśle w styczniu 1982 r ( Flood Caused Majewski, A.(1969), Rozwój hydrograficzny delty by Ice on the Vistula River in January 1982), Wisły w okresie historycznym (Hydrographic Przegląd Geograficzny, 1: 91–111. Development of Vistula River Delta in His- Ingarden, R. (1922), Rzeki i kanały żeglowne w b. toric Times), Przegląd Geofizyczny,14,1: 3–42. trzech zaborach i znaczenie ich gospodarcze dla Makowski, J. (1997), Wały przeciwpowodzio- Polski (Rivers and Navigable Channels in the we Dolnej Wisły, historyczne kształtowanie, Three Former Annexed Territories and their obecny stan i zachowanie w czasie wezbrań Role for Poland’s Economy), Ministerstwo (Embankments of the Lower Wistula, Their Robót Publicznych, Kraków, 670pp. History, Current and Freshet Stage), Hydro- Jacewicz, A.R. and Kuźniar, P. (2000), Zarys technika 24, Instytut Budownictwa Wodnego historyczny kształtowania się doliny i koryta (IBW) PAN, Gdańsk, 427pp. Wisły na odcinku warszawskim (A Historical Makowski, J. and Tomczak, A. (2002), Stany Outline of the Development of the Warsaw wody Wisły w Toruniu w świetle pomiarów Stretch of the Vistula Valley and Channel), z ostatnich dwóch stuleci (Wasserstande der Dokumentacja Geograficzna, Instytut Geo- Weichsel in Thorn auf Grund der Messunges Changes of the Vistula River Channel and Floodplain… 87

aus den letzten zwei Jahrhundert), Studia ers), Gospodarka Wodna, 8: 227–229. Societatis Scientiarum Torunensis, Sectio C, Starkel, L.(ed.), (1990), Evolution of the Vistula v.11,1, Toruń, 174pp. River Valley during the Last 15 000 Years, Manthey, T.and Gilewski, J. (1980), Określenie Part III, Geographical Studies, Special Issue, natężenia transportu rumowiska w postaci 5, 220 pp. unoszonej i wleczonej w ujściowym odcinku Starkel, L., Gregory, K.J. and Thornes, J.B. rzeki Wisły. Studia i Materiały Oceanologicz- (eds.)(1991), Temperate Paleohydrology, John ne, 30, Wiley, Chichester. Maruszczak, H. (1982), Wisła Lubelska in A. Starkel, L.(ed.) (1982, 1987,1990, 1991, 1996), Piskozub, A. (ed.), Wisła. Monografia rzeki. Evolution of the Vistula River Valley during (Monography of the Vistula River), Warszawa, the Late Glacial and Holocene, Part I–VI, 125–136. Geographical Studies, Special Issue 1,4,5,6,8,9, Maruszczak, H. (1984), Spatial and Temporal Instytut Geografii i Przestrzennego Zagospo- Differentiation of Fluvial Sediment Yield in darowania, PAN, Warszawa. the Vistula River Basin, Geographia Poloni- Starkel, L. (2001), Extreme Rainfalls and River ca, 50: 253–270. Floods in Europe during the Last Millenium, Maruszczak, H. (1990), Sediment Transport Geographia Polonica 74,2: 69–79. in the Vistula Drainage Basin. Suspension Szmańda, J. (2000), Litodynamiczny zapis po- Transport. Solute Transport, Geographical wodzi w aluwiach pozakorytowych Wisły, in Studies, Special Issue 5, 85–90. (Lithodynamic Records of Floods in Over- Petts, G.E., Möller H. and Roux A.L.(eds.) bank Alluvia of the Vistula River), Przewodnik (1989), Historical Changes of Large Alluvial z wycieczki terenowej V Zjazdu Geomorfolo- Rivers. Western Europe, J. Wiley , Chichester. gów Polskich, Instytut Geografii, Uniwersytet Plit, J. (2004), Changes in the Middle Course of im. Mikołaja Kopernika (IG UMK), Toruń, the River Vistula in Historical Time, Geo- 221–231. graphia Polonica, 77, 2: 47–61. Trafas, K. (1992), Zmiany biegu Wisły pomiędzy Punzet, J. (1981), Zmiany w przebiegu stanów ujściem Przemszy a Sandomierzem (Changes wody w dorzeczu górnej Wisły na przestrzeni in the Vistula River Course between Przem- 100 lat (1871–1970) (Changes in the Course of sza River Mouth and the Town of Sandomi- Water-Stages in the Upper Vistula Basin over erz), in Trafas, K.(ed.), Wisła w dziejach i kul- 100 years (1871–1970), Folia Geographica, turze Polski, 8, 31–61. ser. Geographia Physica, 14, 5–28. Warowna, J. (2003), Wpływ zabudowy hydrotech- Punzet, J. (1991), Przepływy charakterystyczne nicznej na warunki sedymentacji w korycie (Characteristic Discharges), in Dynowska, powodziowym Wisły na odcinku Zawichost— I. and Maciejewski, M. (ed.) Dorzecze gór- Puławy [The Impact of the Hydrotechnic nej Wisły, cześć I, Państwowe Wydawnictwo Buildings on the Sedimentation in the Vis- Naukowe (PWN), 167–215. tula River Flood Channel in the Reach be- Soja, R. (2002), Hydrologiczne aspekty antro- tween Zawichost and Puławy], Uniwersytet popresji w Karpatach (Hydrological Aspects Marii Curie-Skłodowskiej (UMCS), Lublin, of Anthropopressure in the Polish Carpathi- pp.121. ans), Prace Geograficzne, 186, 130pp. Stachy, J., Fal, B., Dobrzyńska, I. and Hołda- Paper first received: July 2006 kowska, J. (1996), Wezbrania rzek polskich In final form: July 2007 w latach 1951–1990 (Floods of Polish Rivers in the years 1951–1990), Materiały Badawcze, Seria Hydrologia i Oceanologia 20, 80pp. Stachy, J. and Nowak, H. (1977), Przyczynek do zagadnienia zmian reżymu wezbraniowego rzek polskich (Contribution to the Problem of Changes in the Flood Regime of Polish Riv-

BOOK REVIEW

Hungarian Spaces and Places: Patterns of of transition in different contexts, places Transition, edited by G. Barta, É.G. Feke- and spaces. te, I. Kukorelli-Szörényiné, J. Timár, Cen- At the beginning the study deals with tre for Regional Studies Hungarian Acad- regionalisation, an important contemporary emy of Sciences, Pécs, Hungrary, 2005, Hungarian issue. As majority of authors 595 pp.—Dariusz Świątek from the first part of the book suggest, re- gionalisation is an inevitable process, not No more then five years ago the 10th only because of Hungarian presence in inter- anniversary of the transition in central national structures like EU or NATO, and and eastern Europe (CEE) was widely dis- globalization or modernization influences, cussed (Lavigne, 2000), yet the 15th an- but mainly because of national economic niversary went almost unnoticed, without growth. A number of questions arise, such a large number of conferences, special is- as whether new regional structures should sues of journals, reports, books, etc. How- be connected to public administration, how ever it did not past completely unnoticed, resources should be redistributed among now over 15 years since the beginning of the the regions or how to reduce governmental transition from state socialism to capital- control. These questions are answered in ism, some researchers are arguing that this different ways in the first part of the book, process is nearing the end, and some even dependent on the authors’ views. However announced its end. The group of Hungarian what seems to unite almost all of papers is geographers re-assessed 15 years of changes belief that special attention should be paid in Hungary in book Hungarian Spaces and to settlements and regions which are not Places Patterns of Transition. integrated into the network regions. From The book contains five parts which all the chapters in this part of the book, we are created from 33 papers, preceded by a can draw the conclusion, expressed in one of short but very useful introduction on tran- the chapters, that regionalisation still needs sition issues in Hungary. Each paper makes long-term purposeful preparation. one chapter which gives the impression that The processes that have changed Hun- the whole book is rather a compilation than garian economy since the beginning of the a coherent study. Having 33 authors in one nineties re-made also spatial perspectives; book means it is difficult to shape it differ- new differences and inequalities appear in ently. To do justice it should be mentioned different areas and places of public life. The that the four editors made every effort to second part of the book Spatial Processes in arrange the book in the way that makes it the Economy in the Era of Transformation easier to move through the complex prob- presents a number of empirical studies which lems identified by the different authors. deal with major sectors of economy like: ag- Apart from two parts on the direct and riculture, industry, service and retail. The indirect impacts of political, economic and main conclusion that arises from almost all social transition, the book contains three these chapters is that Hungary is experienc- parts on new patterns of spaces, places ing significant and increasing polarisation of and uneven development in Hungary. The economic space. This division is created by analyses of these changes show the effects Budapest and adjacent regions on one hand, 90 Book Review where the majority of industrial investments, contrast with the two last chapters not only financial services or ICT (Information and because they describe transition of rural Communication Technologies) sector are space and small villages, but also because located and on the other hand by rest of the these chapters include lists of suggestions country experiencing considerably lower for rural policy-makers which should lead economic activity. to economic activation in this areas. Transformation also influenced the Euroregions and cross-border coopera- shape of Hungarian social space. New tion are the main topic of the last part of problems appear and others which were the book. All the chapters that deal with very often were hushed up during commu- these issues include an extended historical nism now become more visible. One of the background. After the First World War, the most noticeable and probably unwanted re- country was deprived of 2/3 of its territory sults of changes was creation of unevenness and about 30% of its inhabitants (Kocsis which leads to such phenomena like pov- 1998) and this fact in essential factor in erty, social exclusion or segregation. Social shaping local cross-border cooperation, problems exemplified in the book were os- which seems to be very important in con- cillating around work, gender inequalities, text of European integration. The last two and ethnic problems. All were depicted in chapters of this part raise problems of pro- the chapters grouped in the third (Social tection areas located along country borders trends in transition) and fourth (Changing giving good examples of good cooperation places and spaces) parts of the book. One from Hungarian borders. of the most interesting examples shown Besides the aforementioned lack of here was the emergence of non-govern- analysis of political changes, a picture of mental organisations (NGOs). The analy- the changes in other areas of public life like sis of this phenomenon in the context of education, health care or administration is Hungarian public life brings also questions also missing. However readers will find in- about the consolidation of the democratic teresting in-depth analyses of changes that regime in Hungary. Unfortunately none of take place in Hungarian ‘places and spaces’ book’s chapters is devoted to the policy is- between 1990 and 2005. The majority of sues which seem to be essential in the con- the book has an empirical character, which text of shaping both social and economic helps the reader to understand the nature of reality. changes that took place in Hungary within According its title, Changing places and last 15 years. The book can be recommend- spaces, the fourth part of the book seems to ed not only to readers interested in Hungar- be the core of the volume. This part begins ian changes but also to those interested in with an analysis of employment in Hungary broad transformation processes. (both rural and urban areas) and goes fur- Five years ago Hungarian economist ther to illustrating changes of Hungarian Janos Kornai stated that although econom- urban network in context of access to good ic transformation in Hungary was not over and services, location of knowledge based yet, transition certainly was. The system had investments, innovations and globalisa- become capitalist. What was missing from tion processes. Once again we can observe a successful transformation was a set of strong polarisation but this time of Hungar- broadly- understood institutions that would ian urban space, with Budapest and region- shape contemporary Hungarian space (Ko- al centres on one side and the rest of the rnai 2000). Five year later after reading towns on the other. This impression of po- Hungarian Spaces and Places Patterns of larisation is underlined by two chapters that Transition we can boldly abide by this state- focus on cultural investments and architec- ment. tural changes caused by new global trends in Budapest. All the chapters in this part Book Review 91

REFERENCES: Kornai J., 2000, Ten Years after “The Road to Free Economy”. The Author’s Self-Evaluation, The Lavigne M., 2000, Ten years of transition: a re- World Bank, Annual World Bank Conference view article, Communist and Post-Communist on Development Economics, Washington, Studies, 33, p. 475–483. DC. Kocsis K., Kocsis-Hodosi E., 1998, Ethnic Geog- raphy of the Hungarian Minorities in the Car- Dariusz Świątek pathian Basin, GRI RCES, Budapest. Polish Academy of Sciences Institute of Geography and Spatial Organization E-mail: [email protected]

REFEREES AND ADVISERS TO GEOGRAPHIA POLONICA 2005–2006

The Editors would like to express their sincerest thanks to the referees and advisers who so kindly offered their time to review papers being submitted to Geographia Polonica in the years 2005–2006.

Roy ALEXANDER (Department of Geography and Development Studies, University of Chester, UK); Lars BÄRRING (Department of Physical Geography and Ecosystems Analysis, Lund University, Sweden); Georgy BARTA (Centre for Regional Studies, Hungarian Academy of Sciences, Hungary); Roger BLAIR (Aquatic Monitoring and Bioassessment Branch, Western Ecology Division,United States Environmental Protection Agency (EPA), USA); Niall BURNSIDE (School of the Environment; University of Brighton, UK); Chris J.CASELDINE (School of Geography, Archaeology and Earth Resources, Department of Geography, University of Exeter, UK); Michèle L. CLARKE (School of Geography, University of Nottingham, UK); Paul CLAVAL (UFR de Géographie et Aménagement, Université de Paris IV Sorbonne, France); Louise FILION (Département de Géographie, Université Laval, Québec, Canada); John GORDON (Scottish National Heritage, Edinburgh, UK); Glenn McGREGOR (School of Social Sciences & Public Policy, Department of Geography, King’s College, London University, UK); Jennifer HILL (School of Geography and Environmental Management, Faculty of the Built Environment,University of the West of England (UWE), Bristol, UK); Keith HOGGART (School of Social Sciences & Public Policy, Department of Geography King’s College, London University, UK); Margareta IHSE (Department of Physical Geography and Quaternary Geology, University of Stockholm, Sweden); Sven ILLERIS (Department of Environmental, Social and Spatial Change, Roskilde University, Roskilde, Denmark); Tamara JASNIJUK (Institute of Geography, National Ukrainian Academy of Sciences, Kiev, Ukraine); Martin KENT (The School of Earth, Ocean and Environmental Science, Faculty of Science, University of Plymouth, UK); Hille KOSKELA (Department of Geography, Faculty of Science,University of Helsinki, Finland); Alina KUDINA (Management (Saïd Business School), Oxford University, UK); Tomasz MICKIEWICZ (School of Slavonic and East European Studies, University College London (UCL), UK); Ondřej MULÍČEK (Department of Geography, Faculty of Science, Masaryk University, Brno,The Czech Republik); Malcolm NEWSON (School of Geography, Politics and Sociology, Faculty of Humanities and Social Sciences, Newcastle upon Tyne University, UK); Nina ODING (Leontief Centre, St Petersbourg, Russia); David PASSEMORE (School of Geography, Politics and Sociology, Faculty of Humanities and Social Sciences, University of Newcastle upon Tyne, UK); Michał PASZKOWSKI (Institute of Geography and Spatial Management, Faculty of Biology and Earth Sciences, Jagiellonian University, Cracow, Poland); Vroni RETZEL (Lehrstul Biogeografie, Universität Bayreuth, Germany); Keith RICHARDS (Department of Geography, University of Cambridge, UK); Petr RUMPEL (Department of Geography and Regional Development, University of Ostrava, The Czech Republic); Manuela, Sofia ( Department of Sociology, University of Bucarest, Romania); STANCULESCU Alison STENNING (School of Geography, Politics and Sociology, Faculty of Humanities and Social Sciences, Newcastle upon Tyne University, UK); Juha VÄÄTTÄNEN (University of Helsinki, Finland); Sorina VOICULESCU (Deparetment of Geography, Faculty of Chemistry, Biology and Geography, University of the West Timişoara, Romania); Christian WICHMANN (Department of Geography and Geology, Faculty of Science, MATTHIESSEN University of Copenhagen, Denmark); Denis WHITE (United States Environmental Protection Agency,Western Ecology Division, Corvallis,Or, USA); Paul WHITE (Department of Geography, Faculty of Pure Science, The University of Sheffield, UK); Krzysztof (Institute of Earth Sciences, Department of Environmental Protection, WOJCIECHOWSKI University of Maria Curie-Skłodowska, Lublin, Poland); Piotr WÓJCIK (Chair of Economic Growth Theory, Faculty of Economics, Warsaw University, Poland); Andriej (Leontief Centre; St Petersburg, Russia). ZAOSTROVTSEV