Spatial Aggregation of Mobile Transect Measurements for the Identification of Climatic Microenvironments

Abstract

Mobile transect measurements retrieve high-resolution observations revealing the spatial variation of atmospheric properties throughout an urban landscape. A sensor platform is moved through a study site with varying urban form, collecting a data set that can be used to investigate the multifaceted impacts of different building and landscape configurations on atmospheric properties. To generalize such findings, it is imperative to include transect runs representing different points in time and potentially different meteorological background conditions. However, the analysis of a set of mobile transect measurement runs is challenging because of the strict spatio-temporal dependence and multivariate nature of each recorded sample. In this study, we provide visual support for the identification of coherent climatic microenvironments within a study site using mobile transect measurements taken at different points in time and over diverse routes. A regular grid is used to spatially aggregate the data, and resulting summaries are classified according to similar multivariate relationships using clustering techniques. Finally, each grid cell is visualized using a radial glyph encoding cluster membership, predominant wind direction for each transect run, and the number of transect runs traversing this grid cell. The approach has been tested using a data set recorded in Gilbert, Arizona, USA, and it shows potential to identify spatially contiguous regions of similar microclimate.