PNNL

Abstract

Temporally and spatially explicit population data are vital in the study of many fields, such as demography, urban planning, disaster prevention, economics, and environmental modeling. The population data used in these studies are, however, typically aggregated at census enumeration units. This spatial resolution is too coarse for many applications. Accurate population downscaling methods are needed to obtain population data at finer spatial resolutions. In this study, a novel settlement-related database, Built-Up Property Records (BUPR) from the Historical Settlement Data Compilation for the United States (HISDAC-US), is used to downscale the population from census tracts to block groups. The BUPR dataset, derived from Zillow’s ZTRAX database, provides the number of built-up property records for each 250-m grid at a 5-year temporal resolution from 1810 to 2015 for most of the contiguous United States. The ability of BUPR to accurately downscale populations of four states, representing a range of different geographic regions with different population densities, is evaluated here. BUPR values are used as weights to redistribute the year 2000 population from census tracts to corresponding block groups. The BUPR-based downscaling method is compared with commonly-used methods utilizing other datasets. The BUPR-based method outperforms all other methods in all but one state where the BUPR dataset is incomplete. A more detailed accuracy assessment was also performed by dividing each state into low, medium, and high population density areas. The BUPR method is able to produce more accurate downscaled population estimates for low and medium population density areas when compared with other data sources, though its performance deteriorates in high population density areas due to its relatively coarse spatial resolution. BUPR-based dasymetric mapping was, then, applied to the full, contiguous United States, where it was found to be both highly accurate and generalize well.