Estimate the population spatial distribution of Rio de Janeiro for 2014
ISBN 978-85-88783-11-9
Authors
1Strauch, J.; 2Costa, P.; 3Ajara, C.
1IBGE Email: julia.strauch@ibge.gov.br
2UERJ Email: paulafcosta@live.com
3IBGE Email: cesar.ajara@ibge.gov.br
Abstract
The Rio de Janeiro city in recent years has seen scene of great changes in regional development. These changes are result of city improvements to support the hosting of the World Cup in 2014 and the Olympic Games in 2016. This improvement has attracted labor generating new services and thereby also causing a reordering of urban space with new building and housing projects targeting citizen of class A to class C. These changes have altered the settlement pattern of neighborhoods and their demographic configuration. The population count is performed in Brazil with decadal periodicity and is motivated by the need for demographic information for public policy planning in the medium and long term. Between these demographic censuses there is great uncertainty about the population distribution. However, in some critical applications, for example, to estimate the population affected by any disaster, it is necessary to make estimates of population density. In this scenery there are many proposed methods to estimate population. An important contribution comes from the GIS and remote sensing. In the scientific literature, population estimation methods are divided into two big categories, namely: statistical modeling and zonal interpolation. The statistical regression methods seek to establish a relationship between the population and other variables, especially those derived from remote sensing. The zonal interpolation method is used to reconcile data between different zoning or disaggregate demographic variables. The newer methods are based on using auxiliary information, such as land use and cover map. This paper applies intelligent dasymetric method (Mennis and HUTGREN, 2005) as an alternative method to disaggregate the total population variable in urban area, using auxiliary data in the public domain. So, the 2010 census data and the land use and cover map derived from processing satellite images Landasat 8 acquired in the National Institute for Space Research (INPE) for Orbit 217, point 76, with passage date in 08/10/2014 and central time at 12:52 hours. The image used to classification of land use and cover map is derived from the compositional process 4R; 3G; 2B. After the creation of multispectral imaging image fusion technique is applied using band 8 to integrate the best spatial resolution of the panchromatic band preserving content / color of composite image. The classes of land cover and use map are defined based on the Technical Manual for Land Use (IBGE), adapted to the object of study, being: 1) outcropping of rocks; 2) water body; 3) Forest / Mangrove; 4) Beach; 5) Bare Soil / boulder; and 7) Urban Area / Industrial. The processing of all raster and vector data are performed in ArcGIS 10.1 environment set in UTM projection system, datum SIRGAS-2000 Zone 23 South. The main contribution of this paper is a methodology to estimate population in an urban area using public data and geotechnology.
Keywords
intelligent dasymetric method; estimate population; land use and cover map