Time-space modeling of irregular occupations around Brazilian highways, based on static grids: Case study of BR-408

• Published in: Land use policy 2022-03, Vol.114, p.105971, Article 105971
• Main Authors: de Oliveira Barros, Erison Rosa, Oliveira de Andrade, Maurício, de Souza Júnior, Fernando Lourenço
• Format: Article
• Language: English
• Subjects: Artificial neural networks; Automata theory; Calibration; Case studies; Cellular automata; Dynamic Models; Highways; Irregular Occupation; Land cover; Land use; Mathematical models; Neural networks; Right-of-way; Right-of-ways; Roads & highways; Segregation; Urban development; Urban growth; Urban sprawl
• ISSN: 0264-8377; 1873-5754
• DOI: 10.1016/j.landusepol.2021.105971

Obtaining information about the conditions of the right-of-ways on Brazilian highways represents a primary need in preventing irregular occupations on them. In order to minimize the risks associated with irregular occupation on the right-of-ways, resulting from anthropic processes related to land use and coverage, it is necessary to monitor the areas close to the highways. In this sense, mathematical models based on cellular automata can help to understand the impact of changes in land use and/or coverage and to predict future changes in the margins of highways. The study area comprises the BR-408 highway, located in the forest area of the state of Pernambuco, and aims to assess the urban growth of areas that are crossed by highways, generating an annual land cover scenario for the 10-year period. The modeling carried out in this research used land cover maps from the years 2000 and 2010, obtained through the Mapbiomas Project as input. Eight variables were used to explain the changes in land cover on the margins of Highway BR-408. For the calibration of the Model, the Artificial Neural Network and Logistic Regression method was used. The model was validated by the Kappa concordance test. The results show that the generated scenario presents an increase of 0.50% for the class “Irregular Occupation over the Right-of-Way”, in the period of 10 years. This increase has the contribution of cells that were in the initial stage as class Forest/Non-Forest Formation, Agriculture and non-vegetated area. The results of the modeling point to the growth of occupations in the peripheral areas of the cities cut by the highway under study, showing a spatial segregation. •The behavior of the statistical grid cells in the 5 km buffer located in the right-of-way on the analyzed highway with irregular occupations in 2010 is significantly correlated with population density and housing density.•A time sectioning analysis on interactions of changes on land use classes on neighboring cells (LUCC models) through the MAPBIOMAS data, together with the territory socio-economic profile extracted from census tracts on the margins of the analyzed highway makes it possible to interpret the spatial dynamics of the marginal occupations.•The identification of the socioeconomic profile of marginal territories to highways from disaggregated census data may represent a viable alternative to the need for more accurate data.•Socioeconomic variables and land cover data converted into grids are di