Send to

Choose Destination
Sci Total Environ. 2017 Oct 15;596-597:451-464. doi: 10.1016/j.scitotenv.2017.03.164. Epub 2017 Apr 26.

Determinants of single family residential water use across scales in four western US cities.

Author information

Department of Geography, Portland State University, Portland, OR 97201, United States. Electronic address:
Department of Geography, Portland State University, Portland, OR 97201, United States.
Department of Planning, University of Arizona, Tucson, AZ 85721, United States.
School for the Future of Innovation in Society, Arizona State University, Tempe, AZ 85287 United States.
School of Geographical Sciences and Urban Planning, Arizona State University, Tempe, AZ 85287, United States.


A growing body of literature examines urban water sustainability with increasing evidence that locally-based physical and social spatial interactions contribute to water use. These studies however are based on single-city analysis and often fail to consider whether these interactions occur more generally. We examine a multi-city comparison using a common set of spatially-explicit water, socioeconomic, and biophysical data. We investigate the relative importance of variables for explaining the variations of single family residential (SFR) water uses at Census Block Group (CBG) and Census Tract (CT) scales in four representative western US cities - Austin, Phoenix, Portland, and Salt Lake City, - which cover a wide range of climate and development density. We used both ordinary least squares regression and spatial error regression models to identify the influence of spatial dependence on water use patterns. Our results show that older downtown areas show lower water use than newer suburban areas in all four cities. Tax assessed value and building age are the main determinants of SFR water use across the four cities regardless of the scale. Impervious surface area becomes an important variable for summer water use in all cities, and it is important in all seasons for arid environments such as Phoenix. CT level analysis shows better model predictability than CBG analysis. In all cities, seasons, and spatial scales, spatial error regression models better explain the variations of SFR water use. Such a spatially-varying relationship of urban water consumption provides additional evidence for the need to integrate urban land use planning and municipal water planning.


Single family residential water use; Spatial planning; Spatial regression; Urban water use; Western US

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center