Analysis of Spatial Uncertainty in LiDAR-derived Building Data and Uncertainty Propagation in Modeling of Urban Atmospheric Dispersion

Abstract

Results of environmental models (EMs) are often used to assist decision making. However, EM outcomes vary significantly with different input data, model parameters and model assumptions. Therefore, informed decision making requires an in-depth understanding of how the changes in input data, model parameters and model assumptions influence the model outputs. While EMs are now accustomed to geo-spatial data, the influences of spatial uncertainty are often overlooked. This research examines the influence of spatial uncertainty throughout the three stages of general environment modeling: 1) examine the uncertainty in geo-spatial data as representation of the environment, 2) examine the uncertainty in the linkage between EMs and Geographic Information System (GIS) and, 3) examine and compare the influence of spatial uncertainty with the uncertainty of model parameters. LiDAR data and urban atmospheric dispersion model (UADM) are used as a use case, to demonstrate the methods and benefits of examining the influence of spatial uncertainty toward EMs.