Development and analysis of land-use/land-cover spatio-temporal metrics in urban environmentsexploring urban growth patterns and linkages to socio-economic factors

Resumen

This thesis addresses the development and analysis of new tools and methods for monitoring and characterizing urban growth using geo-data and land-use/land-cover (LULC) databases, as well as exploring their relationships with socio-economic factors, providing new evidences regarding the use of LULC data for urban characterization at different levels by means of spatial and statistical methods. First, the most common spatio-temporal metrics were compiled and implemented within a software tool, IndiFrag. Then, we present a methodology based on spatio-temporal metrics and propose a new index that quantifies the inequality of growth between population and built-up areas to analyze and compare urban growth patterns at different levels. This allowed for a differentiation of growing patterns, besides, the analysis at various levels contributed to a better understanding of such patterns. Second, we quantified the two-way relationship between the urban structure in cities and their socio-economic status by means of spatial metrics issued from a local climate zone map for 31 cities in North Rhine-Westphalia, Germany. Based on these data, we quantified their relationship with socio-economic indicators by means of multiple linear regression models, explaining a significant part of their variability. The proposed method is transferable to other datasets, levels, and regions. Third, we assessed the use of spatio-temporal metrics derived from LULC maps to identify urban growth spatial patterns. We applied LULC change models to simulate different long-term scenarios of urban growth following various spatial patterns on diverse baseline urban forms. Then, we computed spatio-temporal metrics for the simulated scenarios, selected the most explanatory by applying a discriminant analysis and classified the growth patterns using clustering methods. Finally, we identified empirical relationships between socio-economic indicators and their change over time with the spatial structure of the built and natural elements in up to 600 urban areas from 32 countries. We employed random forest regression models and the spatio-temporal metrics were able to explain substantially the variability of socio-economic variables. This confirms that spatial patterns and their change are linked to socio-economic indicators. This work contributes to a better understanding of urban growth patterns and improves knowledge about the relationships between urban spatial structure and socio-economic factors, providing new methods for monitoring and assessing urban sustainability by means of LULC databases, which could be used by researchers, urban planners and decision-makers to ensure the sustainable future of urban environments.