Dissertations and Theses @ UNI
Geospatially-intelligent three-dimensional multivariate methods for multiscale dasymetric mapping of urban population: Application and performance in the Minneapolis-St. Paul metropolitan area
Open Access Thesis
Minneapolis (Minn.)--Population--Mapping; St. Paul (Minn.)--Population--Mapping; Geographic information systems;
The wide availability of remote sensing data, the development of computer technology, and the accessibility of census data in the digital form created new opportunities for highly accurate population estimates. Of particular scientific interest is the method of dasymetric mapping, which can significantly improve the spatial accuracy of mapping socio-demographic processes. In addition to population density, the method has considerable potential in mapping the distribution of other social, economic and demographic variables, such as income level, crime, ethnicity, etc. Another significant gap in the existing studies is the development of three-dimensional dasymetric mapping methods.
This study is focused on developing intelligent dasymetric mapping methods to create algorithms for near real-time display demographic and other socio-economic parameters and assess their accuracy and their potential for geovisual analytics. The study is developed and tested in Minneapolis-Saint Paul area, Minnesota, USA as a key study site given the relative diversity of urban areas and the accessibility for field surveys.
The goal of this study is to develop and test an effective geospatially-intelligent method and GIS algorithm for the creation of multivariable three-dimensional dasymetric (3DM) geographic visualizations for the Twin Cities Metropolitan area. 2D and 3D binary dasymetric mapping methods, as well as floor fraction and intelligent dasymetric mapping method were used to identify the best performing method in terms of accuracy.
The 3D dasymetric mapping method yielded the best accuracy in estimation of population counts in conditions of the given study area. 3D dasymetric mapping method proved to improve the accuracy of population mapping in an urban environment compared to 2D methods. The improvement is more significant at a smaller scale of analysis that reflects a more heterogeneous residential building infrastructure. Finally, the additional socio-economic variables, such as aggregated income and three different types of spending (for food, household supplies, and apparel) were mapped.
The study faced the limitations of the inability to obtain data, perfectly synchronized in time between all the spatial layers, non-straightforward nature of the selection of residential/non-residential buildings and low height variance in the study area.
The future directions of the study are to integrate the developed methods with the existing web mapping platform, test the dasymetric mapping approach on the extended set of socioeconomic variables and explore the usefulness of the dasymetric mapping approach on the smaller scales of the enumeration units and dasymetric mapping polygons.
Year of Submission
Master of Arts
Department of Geography
Andrey N. Petrov, Chair
1 PDF file (vi, 106 pages)
©2021 Nikolay Golosov
Golosov, Nikolay, "Geospatially-intelligent three-dimensional multivariate methods for multiscale dasymetric mapping of urban population: Application and performance in the Minneapolis-St. Paul metropolitan area" (2021). Dissertations and Theses @ UNI. 1106.