Every Building On Earth – The Global Dynamic Exposure Model

The built environment is the largest unknown in understanding the effects of disasters and assessing their risks. It is described by exposure models, which cover structural types of buildings, their vulnerability, occupancy, size and monetary value together with the number of people inside. In classical large-scale models such information is given aggregated per administrative unit as surveying all this information on the building level is virtually impossible. They are not capturing the constant changes of the built environment, through urbanization and modernization of the building stock. Many tasks in disaster risk reduction need up-to-date and high-resolution building information to accurately capture or assess the physical effects of disasters on society and the built environment. Here we introduce the first homogeneously defined open and fully reproducible global exposure model at the building-level resolution derived from open and partially crowd-sourced data. The model, currently containing approx. 2.7 billion buildings, is designed to be dynamic in that it recomputes the exposure regularly as in average 150,000 building datasets are being added per day and thereby accounts for the need of updated exposure data shortly after natural catastrophes. This model allows for capturing the consequences of hazards through exact locations of buildings and their exposure indicators. This is in particular important for localized hazards, such as strong earthquake shaking exacerbated by local site conditions or (flash-)floods affecting a well-defined geographic extent based on elevation and local topography. Our model allows for a deeper understanding of the built environment on a global scale but with a high resolution locally, by assessing each building separately and assessing its properties deterministically, where possible, and probabilistically, where needed. In its core, it is a highly detailed model of buildings, relevant in many fields, such as sustainability research, urban planning or in the insurance industry. Finally it supports targeted preparation and response by disaster managers.