HR: 1330h
AN: U52A-0023    [PDF]
TI: Urban Heat Island Connections to Neighborhood Microclimates in Phoenix, Arizona: Defining the Influences of Land Use and Social Variables on Temperature
AU: * Prashad, L C
EM: lela.prashad@asu.edu
AF: Arizona State University, PO Box 873211, Tempe, AZ 85287-3211
AU: Stefanov, W L
EM: will.stefanov@asu.edu
AF: Arizona State University, PO Box 873211, Tempe, AZ 85287-3211
AU: Brazel, A
EM: abrazel@asu.edu
AF: Arizona State University, PO Box 873211, Tempe, AZ 85287-3211
AU: Harlan, S
EM: Sharon.Harlan@asu.edu
AF: Arizona State University, PO Box 873211, Tempe, AZ 85287-3211
AB: Phoenix, AZ is known to have an urban heat island that significantly increases minimum and maximum temperatures, which continue to climb as the city grows and becomes denser. We present a study that investigates "neighborhood" scale (1 square km) microclimate and its potential connections to the regional heat island. The purpose of our study is to: 1) identify social factors/ behaviors that influence temperature on a neighborhood scale and relate fluctuations to the overall heat island; 2) determine the effect of land use on temperature at the neighborhood and regional scales; 3) evaluate a range of thermal infrared (TIR) remotely sensed (RS) data and compare the RS surface temperatures to air temperature. Neighborhoods in both the urban core and fringe were delineated within Phoenix for our study. The neighborhoods represent a range of income levels and ethnicities. Daytime TIR data from Landsat sensors (TM, ETM+) and the airborne MASTER sensor were used to obtain surface temperatures for the neighborhoods. Nighttime surface temperature data were obtained from the ASTER sensor. Vegetation indices (SAVI) were created from Landsat and MASTER imagery. Climate monitors installed in each neighborhood recorded air temperature and dew point readings every 5 minutes. Land use was obtained from an expert systems classification of Landsat imagery and from aerial photos. Our results indicate surface temperatures correlate strongly with air temperatures. The 12.5m/pixel MASTER and 30m/pixel Landsat thermal data can highlight surface temperature gradients within a neighborhood while nighttime ASTER data provides better mean surface temperature discrimination between neighborhoods, and allows for quantification of local diurnal temperature variation. Neighborhoods with a low mean income, high percentage of Hispanics, and low educational attainment are significantly hotter than their high-income, non-Hispanic, highly educated counterparts. Urban core neighborhoods with high income also correlate strongly with high amounts of vegetation (R= -.637) and have significantly lower surface and air temperatures than regional heat island models predict. This suggests that neighborhoods with the means to alter their environments with vegetation can also produce more amenable microclimates. Conversely, neighborhoods with a high area percentage of concrete, asphalt roadways, and built materials exhibit a strong positive correlation with increased surface and air temperature.
DE: 3309 Climatology (1620)
DE: 3322 Land/atmosphere interactions
DE: 3360 Remote sensing
SC: U
MN: 2003 Fall Meeting