DOI: 10.15201/hungeobull.70.1.1Oşlobanu, C. and Alexe,Hungarian M. Hungarian Geographical Geographical Bulletin Bulletin 70 70 2021 (2021) (1) (1) 3–18. 3–183 Built-up area analysis using Sentinel data in metropolitan areas of Transylvania, Romania Constantin OŞLOBANU and Mircea ALEXE1 Abstract The anthropic and natural elements have become more closely monitored and analysed through the use of remote sensing and GIS applications. In this regard, the study aims to feature a different approach to produce more and more thematic information, focusing on the development of built-up areas. In this paper, multispec- tral images and Synthetic Aperture Radar (SAR) images were the basis of a wide range of proximity analyses. These allow the extraction of data about the distribution of built-up space on the areas with potential for economic and social development. Application of interferometric coherence and supervised classifications have been accomplished on various territories, such as metropolitan areas of the most developed region of Romania, more specifically Transylvania. The results indicate accuracy values, which can reach 94 per cent for multispectral datasets and 93 per cent for SAR datasets. The accuracy of resulted data will reveal a variety of city patterns, depending mainly on local features regarding natural and administrative environments. In this way, a comparison will be made between the accuracy of both datasets to provide an analysis of the manner of built-up areas distribution to assess the expansion of the studied metropolitan areas. Therefore, this study aims to apply well-established methods from the remote sensing field to enhance the information and datasets in some areas lacking recent research. Keywords: backscattering, metropolitan areas, supervised classification, urban footprint, built-up area Received February 2021, accepted March 2021 Introduction Zakeri, H. et al. 2017) to approach the results of high-resolution Synthetic Aperture Radar Current geographic studies attempt to follow (SAR) platforms. Therefore, extensive studies as accurately as possible the different natural were accomplished using different types of and anthropogenic phenomena in the world. classifications (e.g., Corbane, C. et al. 2017) In this direction, different geographic branch- and exploiting Landsat multispectral images es develop techniques for processing and in- and Sentinel SAR images for the global map- terpreting geographic information, such as ping of human settlements, using Global Hu- satellite data. A good example would be the man Settlement Layer, which includes global European Space Agency (ESA) data acquired multi-temporal evolution (1975, 1990, 2000 by the remote sensing satellites, Sentinel-1 and 2014) of built-up surfaces. and Sentinel-2. European satellite data pre- Studying the expansion of the built areas, sents the best performance regarding open- different methods have been applied. One of source multispectral images at a spatial reso- these methods is represented by the normal- lution of 10 m. ESA occupies an important ized difference indices: Normalized Difference position, and its data are being studied and Built-up Index (NDBI) and Normalized analysed by various researchers (Koppel, K. Difference Vegetation Index (NDVI) (Zha, Y. et al. 2015; Khalil, R.Z. and Haque, S.U. 2017; et al. 2003). Then the technique evolved, creat- 1 Faculty of Geography, GeoTomLab, Babeş-Bolyai University, 5–7, Clinicilor Street, 400 006 Cluj-Napoca, Romania. E-mails: [email protected] (corresponding author), [email protected] 4 Oşlobanu, C. and Alexe, M. Hungarian Geographical Bulletin 70 (2021) (1) 3–18. ing new indices based on either the thermal method on Landsat and SPOT multispectral band (As-Syakur, A.R. et al. 2012) or on the images or SAR data (Zoran, M. and Weber, analysis of built-up areas on extended sur- C. 2007). Another example is the comparison faces using a group of built-up indices (Li, H. of Bucharest city with French Guyanese areas, et al. 2017) or combining several vegetation, using the fusion of optical data and SAR data water and built-up indices to reduce confu- with high-resolution (Corbane, C. et al. 2008). sions (Xu, H. 2010). Afterwards, these vali- The MLC is a supervised method, which as- dated indices begin to be used in studies for sumes that the user identifies by visual analy- measurements of the built space (Kaimaris, sis, polygons of pixels or groups of pixels, de- D. and Patias, P. 2016). fining the ranges of spectral values that have Another variant of emphasizing the built-up a correspondent in phenomena or objects in areas is the one in combination with other land the real environment. Then, the classifier de- use classes (Yuan, F. et al. 2005; Dewan, A.M. termines according to statistics which pixel is and Yamaguchi, Y. 2009). In this category, assigned to a certain class that has the highest most of the studies (Sekertekin, A. et al. 2017; probability to be normally distributed. Other Forkour, G. et al. 2018) generated maps using supervised methods (Lillesand, T.M. and supervised classification method (Maximum Kiefer, R.W. 1994) use mean vectors, such as Likelihood Classification, MLC) based on Minimum Distance method and classifies pix- Landsat scenes, then comparing the results els to the nearest class based on Euclidean dis- with Corine Land Cover (CLC). Apart from us- tance or such as Parallelepiped classification ing land cover datasets, other digital resources based on n-dimensional parallelepiped, where may be used for mapping urban areas, such as each pixel is assigned to a certain class defined high-resolution imaging studies, orthophoto by the standard deviation threshold from the maps, the Google Earth data catalogue or even mean of each identified class. The MLC is pre- images acquired by the drones. The approach ferred by some authors for several of regions based on supervised classifications was de- from Romania, such as the Iași Metropolitan veloped even on large surfaces (Ma, Y. and Area (Cîmpianu, C. and Corodescu, E. 2013), Xu, R. 2010) or on long-term models of maps the Brașov Metropolitan Area (Vorovencii, (Padmanaban, R. et al. 2017) using optical data. I. 2017) or the Constanța Metropolitan Area Similar to the trend of the universal sci- (Corodescu, E. and Cîmpianu, C. 2014). Also, entific literature in the field, the tendencies the use of NDVI or NDBI is appropriately ac- from Romania approaches the same remote complished on small human settlements, such sensing elements for studying space in the as Lugoj Municipality and surrounding area extra-atmospheric environment. On this (Copăcean, L. et al. 2015) and for those located subject, a bunch of research concentrated in various natural conditions, in the mountain- on the biggest city, the capital of the state, ous area or the areas with a temperate marine Bucharest. The main trend in the Romanian climate, near the Black Sea lagoons (Huzui, literature was to analyse urban expansion us- A.E. et al. 2012). Besides, optical data is also ing supervised classifications from Landsat used to identify agricultural land conversions scenes and then comparing with CLC data to Argeş County (Kuemmerle, T. et al. 2008). and applying buffers every 5 km to observe Various elements of cartographic represen- the evolution of all elements in the terri- tation methods have been treated in other re- tory (Mihai, B. et al. 2015), but there were search papers, such as cartograms and buff- authors who also relied on high-resolution ers. (EEA, 2006; Grigorescu, I. et al. 2012). It panchromatic and multispectral images, like is in regard to cartogram maps with annual CORONA and IKONOS imagery (Sandric, I. surface growth rates of the built-up area at et al. 2007). In the same period, more complex administrative-territorial unit (ATU) level subjects were applied by other authors, such within the Metropolitan Area of Bucharest as the Principal Component Analysis (PCA) (Grigorescu, I. et al. 2014). For better at- Oşlobanu, C. and Alexe, M. Hungarian Geographical Bulletin 70 (2021) (1) 3–18. 5 tainment of this method, this study will use which can offer a practical comparison be- cartogram maps for comparing the built-up tween two fields of remote sensing, optical/ area percentages resulted from processing multispectral and radar. But the main signa- multispectral datasets and also SAR datasets. ture lies in the proximity analysis with which Regarding the most important studied it was able to notice under what natural con- area, which is in full economic and social ditions the methods succeeded in identify- growth, Cluj-Napoca Metropolitan Area, ing properly built-up areas and which of the this is more intensively studied through the metropolitan areas managed to create exam- perspective of the national university center ples of efficient structures for urban sprawl. present in this city. Thus, in this area, it en- counters various spatial-space studies such as the spatial-temporal expansion of imper- Study area and data meable surfaces using the Landsat data for supervised classifications (Ivan, K. 2015) or Study area the extraction of built-up areas using tex- ture analysis of SAR images combined with Study areas were represented by 6 metro- unsupervised classification Sentinel images politan areas: Baia Mare Metropolitan Area, (Holobâcă, I.H. et al. 2019). Brașov Metropolitan Area, Cluj-Napoca Met- Other studies (Mucsi, L. et al. 2017) are ropolitan Area, Oradea Metropolitan Area, focusing on using more precise instruments Satu Mare Metropolitan Area and Târgu (e.g., hyperspectral aerial image) for accu- Mureş Metropolitan Area (Figure 1). These rate detection of anthropogenic elements. It metropolitan areas were chosen on the basis has reached a level where SAR images are of consulting different scientific articles, tech- increasingly exploited in geographic and nical and scientific reports, development strat- interdisciplinary studies by using interfero- egies, and information from sites managed by metric coherence (Koppel, K. et al. 2015). The metropolitan associations (FZMAUR, 2013).