HTML

ePub

PDF

INTRODUCTION

Dynamics are an important component in analyzing a landscape (Turner, 1989), and, therefore, studying them allows forunderstanding how certain processes evolve with time. A classic example is cited by Dunne and Leopold (1978), who demonstrate the raise in floods occurring in drainage basins after the covering and land usage, as well as drainingnetwork means, have been altered.

There are currently many academic papers that use methodologies which are set in geotechnologies that aim to analyzethe dynamics in the drainage network of a landscape, whether it is a natural or a built one. However, when one looks to analyze these changes within a historical point of view, the available bibliography and methodology arerarer. In this sense, this article aims at analyzing the drainage network dynamics in the genesis area of the city of Petrópolis, in Rio de Janeiro State, comparing the recent spatial arrangement of the drainage networkwith the one in the 19th century, to be more exact on the year 1846, a settlement period in this area.

The city of Petrópolis, in Brazil, is located on the mountain region of Rio de Janeiro State, as has been dealinghistorically with environmental issues with various characteristics. Among them, the mass movements and the reoccurring flood events. This framework is directly linked to the occupation of the city as a whole, but mainly onthe first district, since its first occupation on the 19th century. This city has a different history background from most Brazilian cities, since it was created after an imperial decree and it was defined by a map, the PetrópolisMap (1846) – also known as the Koeler Map – which served as a base for the plan of occupation of the future city. This imperial decree was signed by the Emperor Pedro II of Brazil and was registered in the Administrativebook #155. The decree has allowed for the creation of a settlement and the construction of an imperial house where the old farm Córrego Seco was located, which was acquired by Dom Pedro II’s father, the EmperorDom Pedro I, thirteen years before (Ambrozio, 2012). This decree has become known as “Summer Palace Plan” our Koeler Plan. The plan has allowed for the leasing of the Imperial farm, formerly known as CórregoSeco farm and Corcórdia farm, to Major Júlio Frederico Koeler, a German immigrant who joined the military engineering sector of the Brazilian army. This decree has resulted in a document signed on July 26th,1843 that was a part of the prior, original decree. Ever since the signing of this document, the renter, Júlio Federico Koeler, was obliged to prepare the future map of Petrópolis, dividing the imperial landsinto lots or numbered points which were to be conceded. In addition, he had to plan the Emperor’s Palace and his rooms (Paula Buarque, 1940). The Koeler Map contain different cartographical elements such as hydrography,streets, public parks, land installments in which blocks were divided and plots reserved to public and religious buildings (Sá Earp, 1996; Souza, 2013). It is worth noticing that this map was in an advanced stage ofdegradation and was restored in 2016 (Neves and Zanatta, 2016) (figure 1).

Currently, the municipality is divided into five districts. The city was founded, and its origins were set in the first district, called Petrópolis, where the area covered by the Koeler Map islocated. The area the map covers (of 15.91 square kilometers) represents 12% of the current first district of Petrópolis and 2% of the whole current city. The image 2 shows the perspective of the research study, whichcoincides with the area in the Koeler Map.

An important characteristic of the Koeler Plan is that it can be understood as a project that was worried about theresponsible use of resources and the environment. Rabaço (1985) goes on about important aspects that characterize the Koeler Plan as a project that had a preoccupation with the responsible usage of the environment.This plan is considered the first building law of the city of Petrópolis, forbidding constructions on top of hills and horizontally subdividing lots, that is, making them parallel with the front line. The plan alsoestablished that the front of all houses should face the rivers, and that the sewage should be directed to the pits at the back of the uncultivated lands. This characteristic is seen by Rabaço (1985) as one of the mostoriginal in the Koeler Plan, given that it exempted the building to be constructed in the traditional colonial style in which the houses were erected facing backwards to the rivers, that were functioning as a place for discardingwaste. Finally, the planning of the green areas that would contain avenues filled with trees and gardens with hydrangeas, magnolias and silk floss trees.

The map was projected in a scale of 1:5,000 (Fernandes et. al., 2007) and shows a planned occupation following the drainage network, with the streets and avenues running along the rivers (figure 3). TheKoeler Plan can be seen as an urban plan whose spatial reference element is the Petrópolis Map or Koeler Map (1846), which also functions as a guide for handling and looks to minimize the effects of landslides and flooding.

The map was projected in a scale of 1:5,000 (Fernandes et. al., 2007) and shows a planned occupation following the drainage network, with the streets and avenues running along the rivers (figure3). The Koeler Plan can be seen as an urban plan whose spatial reference element is the Petrópolis Map or Koeler Map (1846), which also functions as a guide for handling and looks to minimize the effects of landslidesand flooding.

In the projections drawn in the Koeler Map (1846), the straight segmentation of the three main rivers that cross thecity was proposed as a facilitating way of occupying the area. Thus, the rivers (Piabanha, Palatino and Quitandinha) suffered modifications on their courses ever sing the upraise of the city, and, along with the occupationof its soil, which modified aspects of its covering, contributing to the alteration of the watershed’s dynamics of the area. Studying the alterations that took place in this period is a fundamental task in order to comprehendthe flooding problem that devastates the city since its formation.

In this sense, the main questions that guide this article are: Were the canals of the drainage network in the genesisarea of Petrópolis modified in their width and sinuosity? Are the registered floods in areas where the canals and soil covering were more modified? While searching for answers to the aforementioned questions, the workmethodology uses an approach that integrated the usage of GIS (the Geographical Information System), supported in both historical and current data, the measurement of physiographical parameters of the drainage network, theuse of satellite images to map the soil covering and data of flooding in the period of 2011 to 2018.

MATERIALS AND METHODS

The article was developed following different steps in which various materials and methods were used, ever since surveyingand structuring cartographical data, images and inventory of floods, to manipulating methods and data structuring, as well as their spatial analysis. These phases are shown as follows.

Data acquirement

In the first phase of the research, the Koeler Map, the digital cartographic base of Petrópolis and Worldview2’s image. The process of georeferencing was carried out directly on the restored version of the Koeler Map (Neves and Zanatta, 2016). This process has led to the collection of 30 inspection points in the digital cartographicbase, which should necessarily be found on the historical map. The calculation of the Standard of Cartographic Accuracy was made according to the reference values on the Decree #89,817 of June 20th, 1984 (BRASIL,1984). Out of the gathered results, the georeferencing has framed the Koeler Map as pertaining to Class A in the scale of 1:50,000, which grants it positioning errors of less than 25 meters. In this process, the cartographicbase in the scale of 1:10,000 was crucial for obtaining the coordinates of known features and, thus, in giving them the same features in the historical map.

Aiming to map the soil covering and to quantifying the level of the edifications in the watershed area, an image ofwas acquired on earth explorer site of USGS, taken with an MSI instrument, by the Sentinel-2 mission. The table 1 shows the characteristics of the satellite bands of Sentinel-2 that were used.

The flood data were obtained from data of four rainfall or gauging stations run by Instituto Estadual do Ambiente (INEA/RJ)and distributed by the three drainage basins that are part of the genesis area of the city (Piabanha, Palatino and Quitandinha). The data encompasses continuous data surveys of limnometric quota from 2011 to 2018.

Manipulation and structuring of data

On the second phase, the Koeler Map was submitted to a vectorization process of the features that reference hydrography.All watercourses were vectorized in the ArcMac 10.4.1 program, however, only the three main canals were reworked in a later step. The information contained in the vectoral file in shapefile format were: the name of the feature,length and sinuosity rate. This information holds the identification of a stream of water through its official name and information on their measuring.

An atmospheric correction of the bands 2, 3, 4 and 8 was carried out so that the image could be used, in QGIS software,through the Semi-Automatic Classification Plugin (SCP) and then, the spectral signatures of many targets of the image were compared as to assure and visually evaluate the result that was obtained in the atmospheric correctionof the bands that were utilized.

The gauging data were analyzed in relation to the quota of stream overflow established by INEA/RJ, allowing for theidentification of all the flooding events to the four stations. These data allowed for the spatial evaluation of the phenomena, identifying the distributing of floods through the three different water basins of interest.

Construction of the analysis model

The measurement of the width and of the sinuosity rate of the canals were used to identify the modifications occurredthrough the courses of the rivers Piabanha, Quitandinha and Palatino. The width was measured through the Euclidean distance between the two margins, associating a constant width value to the whole canal. The figure 4 belowsummarizes this process that was carried out in ArcMap 10.4.1 software, through the Euclidean distance tool. To each margin one matrix file was created, where the pixel value corresponded to the distance to the margin. Inorder to obtain the constant distance between both margins it was necessary to sum the two files (map algebra), cutting the values to the inside of both margins.

The analysis of the stream sinuosity took into account, in turn, the shape dynamic of the main canals of the studiedarea. The shape of the canals can be considered straight, anastomosed, intertwined and meandering, forming the canal pattern (IBGE, 2009). The distinction between the straight, the sinuous, meandering and tortuous reflectsitself in using qualitative and nominal criteria. By its own turn, when using quantitative criteria, one is drawn to problems that involve the adoption of limiting values, as in the proposition of Leopold and Wolman (1957)and of Leopold et al. (1964), indicating that the sinuosity rate is justifiable when it implies another typical element of the kind of canal found in the three river basins. The sinuosity rate is obtained [through] “[…]dividing the length of the canal a in a determined section by the length of this section measured along the valley” (Christofoletti, 1981). The described formula can be equated in .

Where Is is the sinuosity rate, Cc is the length of the main canal and Cv is the vectoral distance between the initialand final points of the canal in the considered section. These values were raised to the main canals of each of the three currently analyzed basins. They were framed in the classification by Leopold and Wolman (1957), thatis: to levels under 1.5, the canal was considered straight; and, to those equal or over 1.5 were attributed the classification of meandering. Starting from the measurement of these two parameters, both in the historical mapas well as in the current cartographic base, the aim is to evaluate modifications occurred on the course of the sections of the analyzed canals.

Soil covering and edification level

With the goal to analyze the urban landscape of Petrópolis’ structure, a mapping o the soil covering wascarried out, aiming to find the classes of “vegetation” and “non- vegetation”. The classification was done on InterIMAGE open source software, which classifies satellite images using an object-orientedclassification technique (GEOBIA), in which the objects of interest are identified and classified in accordance to a series of pre-established parameters, which differentiate them from traditional pixel to pixel techniques,and avoid the “salt-and-pepper” effect (Blaschke et al., 2014). After being classified, areas of both classes were calculated, allowing for the analysis of the landscape modification levels on the river basins.

When using a map of soil covering in the analysis of the draining network in the original occupation area of Petrópolisin the face of the flooding events. Thus, the level of alteration of the soil covering can help us understand in which way do the landscape modifications interfere in the floods that happen in the studied area.

RESULTS AND DISCUTION

The results gathered by this research can be divided in two parts and contribute to reinforcing the importance of HistoricalCartography as a tool for integrated landscape analysis using GIS. The first part is directly about the canals’ width and its sinuosity, both measured in the historical map and in the current cartographic base; the secondpart, on its turn, presents more recent data that follow the direction of observing the impact of landscape transformation as partly responsible for recurring problems of flood in the river basins of the Piabanha, Quitandinhaand Palatino rivers.

The shapes of the canals: sinuosity rate

The application of the aforementioned methodology has allowed for the demonstration of the results about the lengthof the canals, the vectoral extension in the extremes of the rivers section and sinuosity rates in both of the utilized bases. The table 2 presents the values obtained to the three main canals of each analyzed basin.

The Quintandinha river was the one that presented the greater variation in the calculated sinuosity rate, when comparingthe Koeler Map and the cartographic base of the year 1999. The sinuosity rate in this river has diminished from 1,523 to 1,487, which shows a greater rectilinear shape progressing in this section of the canal. This diminishingis a reflex of the modification of length values at the section of the canal (Cc), which has reached the value of 197,99 meters, and of the vectoral length (Cv), much smaller, reached around 11,65 meters. Along with otherresults obtained, this is an important index that helps understanding the flooding problems at the region. For the values obtained in the Palatino river, the length (Cc) has modified in 22,91 meters and raised in 45,25 metersfor the vectoral length (Cv). The considerable difference in the vectoral length can be due to the positioning of the map in a known system of coordinates. Last, the Piabanha river is the one which presented the least variationin the sinuosity rates (+0,575). The length values of the canal section (Cc) have raised 43,98 meters and the vectoral length has had a small variation in the vectoral distance (Cv) of -0,84 meters.

The results of sinuosity rates show that there were alterations to the three main canals of different proportions. Itis necessary to take into account that the Koeler Map holds positioning and precision errors linked to its elaboration and manipulation in a GIS environment. Thus, when thinking about quantitative measurement, the variationscan be greater or smaller than the presented values. However, they are not very significant in the face of the cartographic criteria used to evaluate them, which meet the parameters of the class A of 1:50,000 scale (BRASIL,1984). On the other hand, when taking qualitative aspects into account, when making a comparison, it is possible to note variations on the sinuosity of the rivers, and they thoroughly meet the quantitative data gathered.

The canals shape: canals width

The canals width was measured using the Euclidean distance measuring and, thus, an option was made to show the maximum, minimum and medium results of each of the scenarios. The table 3 presents the results.

The greater variation found in the maximum width of the canal are on the Quitandinha river, with differences of 19,84meters. Along all analyzed canals, comparing both documents used, one can notice the suppressing of features, river islands, as it was mentioned by Fernandes et al. (2017). Another relevant aspect about Quitandinha river isthat the medium value of width in the canal has suffered a variation of only 1,61 m. One can say that the sharpest width variation occurred in an occasional manner and they were concentrated in some sections of the Quitandinhariver. The Piabanha river presents the greater variation in medium width amongst all (7,04 m) and the Palatino the smallest (1,13 m). With respect to the maximum width, its value was 12,38 m and 9,58 m respectively.

Some sections present great discrepancies when the map of 1846 and the cartographic base of 1999. There is a referenceto the existence of river islands in the Koeler Map, where such references have ceased, due to the modification of the rivers course and its rectilinization process. Figures 5 and 6 present four sections where the cited alterationsin respect to the measurements and the suppression of islands are more evident.

The illustration above shows a section of the rivers Palatino and Piabanha with the suppression of river features, inthis case, of islands. In these two rivers the suppression of one island in each river has been identified. It is possible to say that, due to that suppression, theses sections had a larger width in the Koeler Map than inthe 1999 base. This same pattern repeats itself through the canals, even without the suppression of other river features. In figure 5, the caption indicates a 17,81 to 22,48 meters width to the section of the Palatino river,which was not the case through the course of the canal in the 1999 base. In the highlighted section at the Piabanha river, the width values have reached the value of 15,40 to 20,51 meters, and when the same section is analyzedin the cartographical chart of 1999, is possible to realize that the areas of greater width were suppressed, as well as the features of the river island themselves. For the Piabanha river, the highlighted section in the figure5 was represented as a simple line, which, on the mapping scale of the 1999 base, indicates section with a 5 m width at max.

The suppression of two river islands is identifiable, for both sections of the Quitandinha river presented on the figure 6, one in each presented section. To all of the area of analysis, the suppressionof four river islands was identified, two at the Quitandinha river (measuring 847 and 132 square meters), one at the Piabanha river (218 square meters) and another at the Palatino river (195 square meters). The second sectionof the Quitandinha river, presented on the figure 6, was also mapped in the cartographic base of 1999 as a simple line, which hindered direct comparations with the 1846 map. However, through a field survey, it was possibleto verify that the section was modified, holding a smaller than or equal to 5 m width.

When the data of sinuosity rates of the canal sections are confronted, some points are noticeable because of their behavior.The Quitandinha river has presented a diminishing in its length, a reduction on the sinuosity rate and a lot of difference in the value of maximum width measured on the two dates. This allows for the reading that the riverhas had a greater rectilinization and narrowing of the riverbed. The Palatino river presented a similar pattern, with a reduction of the canal’s length and the reduction of the sinuosity rate but showing a low reductionon the maximum width and medium width. This encompasses the same reading of greater rectilinization and narrowing of the riverbed. Last, the Piabanha river has presented the smallest variation on the sinuosity rate, allowingfor the conclusion that it was roughly maintained. On the other hand, it is possible to deduce that the river has suffered few alterations in its sinuosity but had a reduction in the riverbed at the course of the consideredtime span.

Mapping of soil covering and flooding register

The mapping of soil covering aims at calculating the covering for the basins area. From the calculation of classes areas in each basin, one could observe the edification level of each one of them. Theedification level can be understood as the percentage of anthropic areas, or non-vegetation coverings, such as rocky outcrops. It is important to notice that the variable level of change expresses how much of anthropic areais on the basin, therefore, we can thus define how much of this area was altered since the Koeler Map, when there was no effective occupation in the analyzed landscape. In the table 4 are shown the river basins and their respectivepercentage of edification.

Analyzing the table 4, it is noticeable that the Quitandinha river basin holds the biggest level of change in the soilcoverage. This result meets the values that refer to the rivers sinuosity. Thus, one can observe that the Quitandinha river, which presents least sinuosity, is a part of the river basin with the greatest level of change. Thispattern is repeated on the other analyzed rivers and basins. Thus, the Palatino river holds the second smaller sinuosity rate and the second greater level of change in the landscape, and the Piabanha river presents the greatersinuosity rate and the smaller level of change in its basin.

This same behavior happens when the level of change is related to the width and the suppression of islands in the analyzedbasins, creating the conditions to allow for understanding how the flood occurrences are distributed. These floods, identified in the period from 2011 to 2018, where mostly identified in the rainfall and gauging stations locatedon the basin of Quitandinha river (93,1%), followed by the basins of Palatino (5,7%) and Piabanha (1,2%), that is, the floods that are concentrated on the basin with the least sinuosity, greater modification on width, greatersuppression of islands (in quantity and area) and greater level of change. The figure 7 presents the juxtaposition of the usage and soil covering map to the distribution of floods identified in the rainfall and gauging stationsin the analyzed basins. In this figure it is possible to see that the river basin with less area of vegetation is the one to present a greater number of floods.

The genesis area of Petrópolis has developed in a mountain landscape filled with quite steep slopes and valleyssurrounded by mountains bottoms. The usage and covering mappings has left clear that the areas with vegetation are concentrated in these largely steep slopes, and at the valley bottoms the non-vegetation features are spread,pressuring the river canals, which had their courses altered to fit the factory edifications that were building up in Petrópolis, as well as the growing demand for housing (ANTUNES, 2017; AMBROZIO, 2008). Therefore,analyzing the map on the figure 7, one can observe that the area with more flooding is set on the basin with larger levels of change in soil coverage, which, in its own turn, was also where the greater modifications in theriver features have occurred. It is important to notice that these flooding conditions do not only limit themselves to the covering dynamics and river features, since many other elements that constitute the landscape and barerelations to the floods were not addressed. However, the analyzes at hand already greatly suggests the correlation between the analyzed elements in the face of the flooding phenomenon.

CONCLUSIONS

The Historical Cartography has had a growing participation in the studies for understanding the landscape dynamics andin the investigation of phenomena of different natures. Concerning research in Geography, the analysis of the dynamics of landscape is a quite consolidated approach. In so being, Historical Cartography has a great potentialin contributing to this kind of approach, highlighting different aspects of the research. As was presented, using a historical document (Koeler Map, 1846) has helped obtaining a set of data about the study field. Using differenttechniques in an environment of Geographic Information System it was possible to extract, store, manipulate and analyze the dynamic of the sections of the canals in the studied area, with the goal of comprehending aspectsthat have modified in the course of time.

The employed methodology has allowed the presentation of a set of factors that contribute to the flooding problem inthe genesis area of the city of Petrópolis, whose roots assent not only on current actions, but also on historical ones. The calculation of the sinuosity rate and of the width of the riverbeds has allowed us to compareand analyze the shape of the canals through the considered time period.

Answering the first posed question, the sinuosity of the canals has decreased, indicating a greater straight sectionof the canals. The width of the riverbed is also another measured aspect that can contribute to understanding the number of floods in certain sections. All investigated sections have suffered narrowing in different dimensionsand, in some locations where river islands were once observed, it was possible to see that they were suppressed. Together with the first set of analyses, (sinuosity and width of the canals), the classification of the soilcovering, and the data obtained in the rainfall and gauging stations installed at the canals have all allowed for the interpretation that number of floods in the considered rivers is due to different associated factors.

We could point out signs that the Quitandinha river section, which suffered the greater alteration in sinuosity andwidth, has lost the river islands as well as suffered the most degrading of the vegetation area, is the one that registers the greater number of floods. In an opposite direction, the section of the canal of the Piabanha riverregisters the smallest number of foods and holds the smallest alteration levels in the sinuosity rates and rivers width, as well as in the vegetation area of the river basin. This demonstrates that, in answer to the secondraised question, that the floods occur in areas where there is greater modification of the shape of the canals and soil covering.

Based on the previous considerations, we can conclude that the article was able to show the participation of differentlandscape modification factors, which help in causing floods in the genesis area of Petrópolis. With the aid of historical maps and other analyses in GIS, it was possible to point out to signs that link the registeredfloods and the canals’ shape modifications and soil covering.

AKNOWLEDGEMENTS

This article was carried out with the support of the institution Coordenação de Aperfeiçoamentode Pessoal de Nível Superior – Brasil (CAPES) of the Conselho Nacional de Desenvolvimento Científico e Tecnológico – CNPQ.

BIBLIOGRAPHY

AMBROZIO, J. C. G. O Presente e o Passado no Processo Urbano da Cidade de Petrópolis (Uma História Territorial). Tese de Doutorado, Universidade de São Paulo – USP, SãoPaulo, 2008.

AMBROZIO, J. C. G. O território da enfteuse e a Cidade de Petrópolis – RJ, Brasil. Revista Electrónica de Geografía y Ciencias Sociales, vol. XVI, 418 (39), p.1-7,2012

ANTUNES, F. S. Geotecnologias e cartografia histórica no auxílio à análise da organização espacial da área gênese de Petrópolis - RJ.Dissertação de mestrado, Universidade Federal do Rio de Janeiro – UFRJ, Rio de Janeiro, 2017.

BLASCHKE, T.; HAY, G.J.; KELLY, M.; LANG, S.; HOFMANN, P.; ADDINK, E.; FEITOSA, R.Q.; MEER, F.V.D.; WERFF, H.V.D.; COILLIE, F.V.; TIEDE, D. Geographic object-based image analysis–towards anew paradigm. ISPRS journal of photogrammetry and remote sensing. Vol. 87, pp. 180-191, 2014.

BRASIL. Decreto n. 89.817, de 20 de jun. de 1984. Instruções Reguladoras das Normas Técnicas da Cartografia Nacional. Diário Oficial da União, Seção1, 22/6/1984, 1984, Página 8884.

CHRISTOFOLETTI, A. Geomorfologia Fluvial. Edgard Blucher. São Paulo, 1981.

DUNNE, T.; LEOPOLD, L. B. Water in Environmental Planning. W. H. Freeman. New York, 1978.

LEOPOLD, L.B. AND WOLMAN, M.G. River Channel Patterns, Braided, Meandering and Straight. U.S. Geol. Surv. Paper. 1957, 282-B.

LEOPOLD, L.B., WOLMAN, M.G. AND MILLER, J.P. Fluvial Processes in Geomorphology. Freeman, San Francisco, 1964, 522 p.

FERNANDES, M.C., LAETA, T., SANTOS, D.F. AND MENEZES, P.M.L. Cartographic Memory Preservation of the Petrópolis City in Brazil: Koeler Map Scanning Using Photographic Survey. In: LectureNotes in Geoinformation and Cartography.1 ed.: Springer International Publishing, 2017, p. 3-19.

IBGE - Fundação IBGE. Manual técnico de geomorfologia. Coordenação de Recursos Naturais e Estudos Ambientais. Série Manuais Técnicos em Geociênciasn.5. IBGE. Rio de Janeiro, 2009.

NEVES, F.M.B; ZANATTA, E.M. Traços de Koeler: a origem de Petrópolis a partir da Planta de 1846. Globomídia Comunicação, 1ª Ed., 235 p., 2016.

PAULA BUARQUE, A. História e Historiographos da Cidade de Petrópolis. Rio de Janeiro: O Livro Vermelho dos Telefones, 1940.

RABAÇO, H.J. (1985). História de Petrópolis. Petrópolis. Instituto Histórico de Petrópolis - IHP, 140 p., 1985.

SÁ EARP, F. Um pouco além de Thomas Kuhn. Da história do pensamento econômico à história das ideias econômicas. Revista de Economia Política,16(1), jan/dez, 1996.

SOUZA, B.C.P. A Toponímia como uma forma de resgatar a memória do lugar. Anais do V Simpósio Luso-Brasileiro de Cartografia Histórica, Petrópolis, 2013.

TURNER, M. G. Landscape Ecology: The effect of Pattern on Process. Annual Review of Ecology, 20(1), 1989, 171–197 p.

Author notes

IR