Now that more than half of Earth’s residents live in cities, understanding the timing, location and magnitude of water uses at sub-city scales is becoming more important for managing water resources and planning the associated infrastructure. Urban water use is not spatially or temporally uniform; certain uses dominate at different times and places. Manufacturing, for example, is a concentrated but constant water use, while irrigation is a dispersed but seasonally varying water use. Such differences affect how water systems operate in terms of energy efficiency, water quality and hydraulic performance.
While state and federal agencies regularly collect, analyze and publish water-use data, few efforts have penetrated to sub-city detail. The U.S. Geological Survey, for example, compiles state and county-level data every five years as part of its National Water-Use Information Program. At the city or utility level, water suppliers measure water consumption by address, but usually only for billing purposes; such data are not usually published and are rarely the subject of spatial analysis. To date, published studies on the spatial variation of water use in a given city have been limited to Portland, Phoenix and three cities in Nebraska.
This case study describes a spatiotemporal analysis of water consumption in Salt Lake City, Utah. The analysis includes monthly water use by census tract for the year 2010.
A GIS Approach
Data for this study originated from the State of Utah Open Data Catalog, an online public database launched last year. The tabular data include monthly water use by type (apartment, business, industry, etc.) and census tract (a geographical unit used for demographic analysis). The period of record extends from January 2000 to May 2015, from which the year 2010 was selected for this study.
GIS was the primary tool for this analysis, aided by Python scripts for data processing. First, the dataset for the entire period of record was read and organized into monthly files. The data were then summarized by month and tract. For the selected year, the tabular data were joined to a shapefile of tracts. Finally, data for the selected year were displayed with graduated colors according to water use.
A reference map of the Salt Lake City area, with census tracts overlain, is provided below. Areas numbered in the figure represent distinct land uses that will be related to water use later.
Salt Lake City reference map. Note differing land uses with heavy industry (1), a university campus (2), mixed use (3), residences and golf course (4), and low population (5).
The video below shows the results. Each polygon is a census tract corresponding to the reference map, and each hue represents a range of water use, increasing from light to dark.
Examining the results, several patterns are apparent that confirm what one might expect about the relationship between land use and water use. In general, the citywide pattern of darker hues during summer months (July–September) indicates higher water use, which is attributed to landscape irrigation required in Utah’s semiarid climate.
The dark sliver (1) at top center is a heavy-industrial area with relatively constant water use. The area includes a refinery, railyard and other mechanical and chemical industries.
The large semi-triangular patch at middle right (2) includes a university campus, research park and hospital. Water use is relatively constant, though a summer increase for irrigation is noticeable. The lowest water use occurs in December, which may be attributed to the holiday break when the campus is mostly vacant.
One mixed-use area (3) includes warehouses, light industry, offices, apartments and single-family homes. In this particular year, their combined effect was high water use from January to September, followed by three months of very low water use.
A residential neighborhood and golf course (4) display significant seasonal variation due to irrigation demands. Water use here is greatest in the summer and lowest in the winter.
At left (5) are large but sparsely populated areas that include light industry and an airport. Water use is relatively low and consistent despite the large land area.
Further study with this dataset could include comparison with other years, comparison with other cities, regression with climate variables, spatial statistics and an analysis of water use types within and among tracts. The results may also be input into extended-period hydraulic simulations in order to more accurately represent demand in the city’s water distribution system.
Beyond Salt Lake City
Salt Lake City’s water use varies in space and time. Though such relationships are complex, the techniques presented here provide insight to specific spatial and temporal patterns that can inform water resources management.
A similar analysis may be performed for any city or water district, provided that the data describe the time, location and magnitude of each water use observation. The maps shown here are one example of how the results may be visualized. As an alternative to census tracts, water use may be geocoded by individual address and represented by graduated symbols or graduated colors. Other data, such as water quality parameters, may be mapped by similar methods.
With a compatible historical dataset, this type of analysis can be automated and repeated for any time period or variable.