Modelling Patches Number of Residential Areas in Relation to Some Environmental Characteristics (Case Study: Tehran City)

Abstract

With human intervention, the face of the land is divided into smaller spots and after a while many problems will arise. If the number of patches is less and the land uses are more uniform, the appearance of the land will be less damaged. Therefore, the landscape metrics are indicators of environmental sensitivity. Today, Tehran, as a metropolis, is facing many problems that does not have the physical, spatial and functional structure appropriate to its needs and it has a long distance with sustainability indicators and biological standards. The present study considers the spatial modelling patches number of residential areas in relation to some environmental characteristics as a research necessity for the Tehran city and addressed it. In order to conduct research, some of the most important environmental factors influencing the creation and development of residential areas in Tehran were considered. These factors include: elevation, slope, vegetation, density of rivers, of ground water depth, wind speed and precipitation in the region. Then, a raster layer was prepared for each of the factors in Arc map software. Landsat 8 satellite images were used to prepare the layer of residential areas. In the next step, a grid with 20 rows and columns was placed on Tehran, and the entire area was covered by 238 pixels. For modeling, patches number of residential areas were considered as the response variable and other factors were considered as independent variables. In order to perform negative binomial model, the collected data were analyzed by RStudio software. The results showed that the elevation, vegetation and slope respectively have a positive, positive and negative effect at the 99% confidence level on the response variable (patches number of residential areas) they have. Mapping the patches number of residential areas based on modeling results showed that the number of patches is more in the northeast part of Tehran. This result shows that the continuity of ecological zones is low in these parts. The correlation between the raster layer of the number of real patches and the modeling was 0.856, which shows that the modeling is very accurate and can be used in urban planning.