Resource information
Critical to effective urban climate adaptation is a clearer understanding of the sensitivities of resource demand to changing climatic conditions and land cover situations. We used Bayesian Maximum Entropy (BME) stochastic procedures to estimate temperature and precipitation at the very small scale of urban Census Block Groups (CBGs) in Phoenix, Arizona and Portland, Oregon, and then compared average household water use patterns by climate conditions and land cover characteristics between and within the two cities. Summer household water use was positively related to maximum temperatures and dense vegetation cover in terms of grass cover and trees and shrubs; it was negatively related to precipitation amounts in both cities. Water use was more sensitive to maximum temperature, precipitation levels, and vegetation cover in Phoenix than in Portland. There was substantial intra-city variation with greater sensitivity in urban water use associated with higher densities of trees and shrubs in both cities, but in Phoenix, the highest sensitivities to maximum temperatures occurred in CBGs with the most grass cover while in Portland, high sensitivity was associated with CBGs with the least grass cover. Many of the latter are in highly built-up downtown areas of Portland where artificial irrigation is required to maintain landscapes during the hot summer season. Take-home messages are: (1) BME space/time statistics provide efficient estimates of missing precipitation and temperature data to create continuous high resolution meteorological data that improve water demand analysis and (2) use of landscaping for urban climate adaptation will have differing impacts on water use, depending on local climate conditions, urban layout, and the type of vegetation cover.
