Mortality analysis of an urban building stock

Research is presented on the estimation of the lifespan of cohorts of buildings and building stocks. This is based on the analysis of extensive longitudinal data of the 55 000 buildings in the City of Zurich from 1832 and 2010. The survival probability from different perspectives considers age, construction periods, and demolition periods for both existing and demolished buildings. Survival probability is established using a Kaplan-Meier estimator. A more in-depth approach to the mortality of buildings is then determined by differentiating building age, use, size and geographical situation (district). The use of a common geographical information system (GIS) allows longitudinal building data to be linked to a geographical hierarchy of three levels of analysis (city, district and building) accounting for the different granularity on each level (neighbourhoods, quarters, building parts). A comparison of the two methods indicates that the choice of the observed time periods can lead to very different results. The analysis of the three levels shows the possibilities and limits of combined statistical and historical approaches. Mortality analysis is a promising approach to inform policy and practice; it could become a new link between long-term scenario planning, construction policies and institutional regimes.;Research is presented on the estimation of the lifespan of cohorts of buildings and building stocks. This is based on the analysis of extensive longitudinal data of the 55 000 buildings in the City of Zurich from 1832 and 2010. The survival probability from different perspectives considers age, construction periods, and demolition periods for both existing and demolished buildings. Survival probability is established using a Kaplan-Meier estimator. A more in-depth approach to the mortality of buildings is then determined by differentiating building age, use, size and geographical situation (district). The use of a common geographical information system (GIS) allows longitudinal building data to be linked to a geographical hierarchy of three levels of analysis (city, district and building) accounting for the different granularity on each level (neighbourhoods, quarters, building parts). A comparison of the two methods indicates that the choice of the observed time periods can lead to very different results. The analysis of the three levels shows the possibilities and limits of combined statistical and historical approaches. Mortality analysis is a promising approach to inform policy and practice; it could become a new link between long-term scenario planning, construction policies and institutional regimes.;Research is presented on the estimation of the lifespan of cohorts of buildings and building stocks. This is based on the analysis of extensive longitudinal data of the 55000 buildings in the City of Zurich from 1832 and 2010. The survival probability from different perspectives considers age, construction periods, and demolition periods for both existing and demolished buildings. Survival probability is established using a Kaplan-Meier estimator. A more in-depth approach to the mortality of buildings is then determined by differentiating building age, use, size and geographical situation (district). The use of a common geographical information system (GIS) allows longitudinal building data to be linked to a geographical hierarchy of three levels of analysis (city, district and building) accounting for the different granularity on each level (neighbourhoods, quarters, building parts). A comparison of the two methods indicates that the choice of the observed time periods can lead to very different results. The analysis of the three levels shows the possibilities and limits of combined statistical and historical approaches. Mortality analysis is a promising approach to inform policy and practice; it could become a new link between long-term scenario planning, construction policies and institutional regimes.;