Avaliação de estimativas populacionais com dados de luz noturna VIIRS DNB

• Autor(a) principal: Xavier, Rana Barbara Silveira
• Data de Publicação: 2022
• Tipo de documento: Trabalho de conclusão de curso
• Idioma: por
• Título da fonte: Repositório Institucional da UFSCAR
• Texto Completo: https://repositorio.ufscar.br/handle/ufscar/16607
• topic: VIIRS; Luminosidade; Sensoriamento remoto; Estimativas populacionais; CIENCIAS EXATAS E DA TERRA::GEOCIENCIAS

The population projections of the Brazilian Institute of Geography and Statistics (IBGE) serve as a basis for urban studies, planning and monitoring. In this sense it is essential that they get as close as possible to reality, which is constantly undergoing changes. For this, periodic reviews of the methods used to maintain data quality are necessary. On the other hand, these constant urban changes highlight the need for new data sources in addition to conventional population projections, and nighttime remote sensing demonstrates the potential to meet this demand. Using the Visible Infrared/ Imaging Radiometer Suite (VIIRS) Day/Night Band (DNB) instrument operated by NASA's Defense Meteorological Satellite Program (DMSP) program, it is possible to use the product generated by it, the VIIRS Nighttime Lights (VNL), in different ways, and are generally applied in urban studies in conjunction with land surface and population census data. In view of this, this work proposes to carry out an analysis of the relationship between night light compositions of the VIIRS instrument and the population estimates of the IBGE in the period from 2014 to 2020. On this sense, the methodology applied involved the use of data from population estimates together with the VNL compositions, producing OLS linear regression-scatter plots of the relationship between the population estimate and the luminosity for each year in Brazil, and WLS weighted linear regression for the regional relationship between the population estimates and luminosity, considering the areas as a weighting factor, seeking to identify whether the relationship between the data remains stable over the years. The results suggest a reduction in the relationship over time, which may be affected, among other factors, (i) by the expansion of illuminated (rural) areas; (ii) the impacts of the pandemic on luminosity; as well as (iii) by the possible distance of the estimates in relation to the real population totals. This analysis sought to provide subsidies for the adoption of this type of data as a complementary parameter in population estimates.