GIS: Tirana, Albania 

GIS: Tirana, Albania
Urban Morphology

Tirana, Albania has undergone significant urban change in the past 15 years. Since 2015, however, there has been an increase in building activity. New roads are being built, the Lana River rerouted, and larger buildings are starting to dominate the landscape. In 2015 Stefano Boeri Architetti won a competition to redesign the city with a project Tirana 2030. The project looks into the ways the city has changed since the project was implemented. Using Sentinel-2 satellite false-color infrared imagery buildings and developments are classified to find the changes in the urban fabric. Using this data, the project looks at urban growth, building typologies and patterns, infrastructure improvements, demolition along the Lana River, and resettlement of Roma near the Sharra landfill.

View the full presentation and methodology paper below.


The project looks at the different ways Tirana, Albania has undergone significant urban change. Under Edi Rama, the current Prime Minister of Albania and former Mayor of Tirana, the city has undergone several infrastructure modernizations and urban redesign attempts. In 2009, Albania applied for EU membership. Since 2011 and 2015, under various mayors, international design competitions have been held. Most recently, Stefano Boeri Architetti won the competition with their Tirana 2030 plan[1]. The project looks at how Stefano Boeri's plan has impacted the urban environment. As a 'modernization' project in a 'developing' city, the carving and rebuilding of the urban fabric mean the displacement and relocation of households. This is especially true for those most vulnerable, especially the Roma population. I will look at three aspects of the city's changing urban fabric, comparing different parameters to understand the impact of the changing landscape. Areas of the city I am analyzing correspond to different areas of development.

Tirane Municipality

Tirane is essentially the downtown core of Tirana. At the center of Tirane is Skanderbeg Square, designed by 51N4E architects and built-in 2008[2]. In the 1990's Skanderbeg was an informal marketplace, where small permanent kiosks were scattered throughout the area. The kiosks that dotted the square were illegally built[3]. In 2008, Edi Rama, the then-mayor of Tirana, held a competition to redesign the central square. The construction of today's Skanderbeg Square required the demolition of the kiosks and the conversion of the informal market to a more considerable public development, surrounded by government buildings and high-end residential developments. Smaller individual structures clustered together make up most of Tirana's urban fabric. As the city 'modernizes,' so to does the housing typology.

Tirane has also been the center of infrastructural improvements, such as river rerouting and new roads. As part of the Tirana 2030 plan, the city's central axis has been reestablished by extending Zogu I Blvd and creating New Boulevard Tirana to the north of the city. The area used to be vacant land around old railroad tracks, and now it is an 80' wide boulevard flanked by new and significant developments. Areas with high-intensity development are easily seen on the Sentinel-2 imagery. At times, like the boulevard, the development takes place on relatively vacant land; at others, it requires destroying parts of the urban fabric.

The Lana River

The Lana River is a relatively small tributary to the Tirana River, and it runs directly through the downtown core of the city. The river is a central focus of the new urban plan for Tirana. Since 2015 areas west and east of the central Skanderbeg square have undergone substantial development. Once winding through the city, the river has been straightened and is now a continuous linear park. What is the impact of rerouting the river? To the west of downtown, the river has undergone the most development. Many buildings have been demolished, and the river straightened. Many of the smaller structures that fit into the various turns of the river have been destroyed.

To the west, the Lana had already been developed into a park. Since 2016, several industrial buildings and vacant lots adjacent to the river have been converted into larger residential complexes.

As the city plans to reroute the river in other city areas, how will this impact those living and making a livelihood? The pockets formed by the meandering river were once spaces for junkyards, auto shops, and informal housing. By formalizing the river and increasing development, both jobs and affordable housing are being pushed to the city's fringes.

Sharra Landfill

Sharra is a new landfill, planned in 2008 and brought online in 2009[4]. Since then, there has been a steady increase in garbage brought to the location every year. Before the formalized landfill, there was a waste dump where many Roma settlements had formed around the site. The Roma collect scrap metal and other valuables as a means of income. The settlements were demolished when the dump was formalized into a municipal landfill. Have the Roma community returned to settle around the landfill? As Sharre, the town near the site is relatively rural, the 10-meter resolution of the Sentinel-2 imagery allows for any development to be identified.


I collected Sentinel-2 satellite imagery of Tirana. Bands 8, 4, 3, and 2 allowed me to create natural color and false-color infrared images. I merged bands 8, 4 and 3 to create a false-color infrared visualization helping me distinguish planted surfaces from unplanted ones, and to identify buildings in the city. The city consists of houses surrounded by vacant, farm, or overgrown land. This band combination also helped identify areas with recent development as they were typically areas with an excess of unplanted areas. The construction of New Boulevard Tirana and the rerouting of the Lana River are easily visible in the imagery. The district of Dajt, to the west of the downtown core, was not fully covered in the raster imagery. I decided to keep Dajt as the areas around Tiranë that I had imagery for were highly developed compared to the rest of Dajt.

Before processing the imagery, to maintain a workable file size, I wanted to clip the imagery to the municipalities I would analyze. At the center of Tirana is the district of Tiranë. To get a better understanding of the morphology of Albania's capital, I decided to clip the rasters to Tiranë and its adjoining districts: Berxulle, Dajt, Farke, Kamez, Kashar, Paskuqan, and Vaqarr. I took a shapefile of the selected districts and dissolved it, giving me the outline of the districts and a shapefile that I used to clip the raster.

With the rasters clipped, I ran a classification using dzetsaka's k-nearest neighbor. Three classification levels were used, high-level development, low-level development, and land. High development or high density allows new construction and changes in building typology to be identified between the two years.

In order to find the areas that had changed between 2016 and 2019, I flattened each classification into binary results (land and development) using reclassify by table. I then ran the raster calculator subtracting 2019 from 2016, and it gave me the output of -1 = areas that have become land, 0 = areas that have not changed, and 1 = areas moving from land to development.

I polygonized the results and dissolved by field, giving me three separate polygons. Then, I calculated the area in km2 using the field calculator and created a column "area_km" using the formula $area*0.000001, as the original units were in m2.

With city-wide growth rates, I intersected the results with each district. From this, I could calculate the growth rate on a smaller scale.

To look at specific growth, I intersected the original polygonized raster with the shapefile for Tiranë. To find larger areas of growth, I searched by expression for clusters of newly built areas larger than 3000 m2 (if "area" < 3000, "DN" 0, "DN" -1). 3000 m2 is about the size of a large building, an infrequent typology beyond the city's center. I then dissolved by field to isolate all larger clusters of cells.

A question that kept coming to my mind was, how can I find the areas where new buildings replaced old buildings. Again, I wanted to look in the city center where most high development occurred. I went back to the original classifications with three separate classes. The steps to do that were as follows:

  1. Reclassify 2016 to land and development, keep 2019 with three classes.
  2. Perform the raster calculation "2019 raster – 2016 reclassified raster", identifying areas where high development occurred over existing buildings.
  3. Reclassify the new raster to the binary: land and rebuilt (high development over existing buildings)
  4. Reclassify the original 2016 raster to the binary: land and high development, isolating any previously existing high development, resulting in a 0 if a high development area from 2016 is also identified in 2019.
  5. Perform the raster calculation "land/rebuilt – 2016 high dev".

Using these calculations, I produced a raster that identified three classifications, land, demolished area, and increase in density.

I then turned to the Lana River, the site for increased development and urban change. Knowing that turning the river into an urban linear park has meant demolishing many buildings. To answer the question, "how likely is it for a building along the Lana to be demolished?" I took the shapefile for the river from OpenStreetMap, created a buffer of 50 m on either side (100 m wide area total) that ran down the length of the river. From there, I intersected polygonized reclassified raster data (0 = demolished, 1 = other) with the buffer area. I was then able to retrieve the total area and demolished area. I used this percentage and compared it with the city-wide percentage calculated from the polygonized reclassified raster data.

To look at the change in the built environment around Sharra Landfill, I intersected the original polygonized layer of -1 = areas that have become land, 0 = areas that have not changed, and 1 = areas moving from land to development, with the outline of the Vaqarr District. I then placed the outline of the landfill from OpenStreetMaps and applied a 500 m buffer around the outside. I intersected the buffer with the polygonized raster and then used 'difference' to remove the landfill area.

In order to look for small settlements < 30 m I searched by expression (if “area” < 30, “DN” 0, “DN” -1). I then dissolved by DN to get the binary of 0 = developed areas under 30 m, and -1 = all other areas. From this, I could compare the percent of area developed between 2016 and 2019 that could be informal or illegal. By looking at the spatial pattern of the nodes, a lack of planning logic can indicate informal housing.

I decided to overlay the findings of Sharra Landfill with the ESRI Satellite Imagery to virtually 'ground truth' my findings. 


The imagery resolution allowed me to perform a relatively accurate classification; however, the difference in imagery between 2016 and 2019 proved difficult. By 'ground truthing' using Google Earth, I found that, in a high number of cases, the change was an inaccuracy from the classification and did not exist. Besides classification concerns, significant scale changes in the urban fabric, such as the rerouting of the Lana River, are highly evident; their impact on the change of the built environment is quantifiable through classification. Though there were inaccuracies, I thought that the resulting calculations still provided relevant insight and visual clues into the changing urban landscape of Tirana.

Works Cited

Stefano Boeri Architetti. "tirana 2030: general local plan by stefano boeri has been approved." 04 Jan 2017. Designboom. < 2017/>.

51N4E. Skanderbeg Square.<>.

Pojani, Dorina. "Urban design, ideology, and power: use of the central square in Tirana during one century of political transformations." Planning Perspectives 30.1 (2014): 67-94.

Alcani, Majlinda, et al. “SOME ISSUES OF MUNICIPAL SOLID WASTE MANAGEMENT IN ALBANIA AND ESPECIALLY IN TIRANA CITY.” International Journal of Science Technology & Management, vol. 04, no. 01, Feb. 2015, pp. 445–457.

Data Sources

Raster Data: Sentinel-2, Esa

Albania/Tirana Borders: Humanitarian Data Exchange, Tirana Open Data

Europe Shapefile: Natural Earth Data

Roads, Rivers and Boundaries: OpenStreetMap

Satellite Imagery: Esri Satellite