A MULTI-SENSOR ASSESSMENT OF URBAN HEAT-ISLANDS: 
A CASE STUDY 
Kevin P. Gallo, NOAA/NESDIS/Office of Research and Applications, 
Washington, D.C., 20233 USA 
Timothy W. Owen, NOAA/NESDIS/National Climatic Data Center, 
Asheville, NC, 28801 USA 
Kimberly E. Baugh, Cooperative Institute for Research in Environmental Sciences, 
University of Colorado, Boulder, 
CO 80303, USA 
Christopher D. Elvidge, NOAA National Geophysical Data Center, 
Boulder, CO 80303, USA 
Commission VII, Working Group 5 
KEY WORDS: Remote Sensing, Satellite, Urbanization, Climatology 
ABSTRACT 
Data acquired during the early to mid-1990s by several satellite-sensor systems were combined in an assessment of the urban heat-island 
effect for the Dallas-Fort Worth, TX region of the United States. Normalized difference vegetation index and radiant surface temperature 
were computed from NOAA-AVHRR data. Two measures of the anthropogenic light emitted by urban-related surface features 
were available from the DMSP-OLS. Landsat MSS data were used to provide estimates of the predominant land cover within the 
grid cells associated with the NOAA-AVHRR and DMSP-OLS data. The multi-sensor analysis of the environment associated with 
seven climate observation stations in the Dallas-Ft. Worth region provided a methodology for characterization of the stations as 
“urban” or “rural.” Three of the seven stations examined were identified through this analyis as “urban.” The information provided 
by a single sensor, while valuable, was clearly enhanced by the use of the multiple sensors included in this study. 
INTRODUCTION 
Climatologists have long been interested in the differences in 
observed ambient air temperature between cities and their 
surrounding rural regions, which have been well documented 
(e.g., Landsberg, 1981; Kukla et al. 1986; Karl et al. 1988; 
Changnon 1992; Gallo et al. 1993a; Gallo et al. 1996). Urban 
climate studies have traditionally focused on the magnitude of 
such differences, which collectively describe the Urban Heat 
Island (UHI) effect. This effect is not restricted to large 
metropolitan areas. In fact, it has been detected in cities with 
populations of less than 10,000 (Karl et al., 1988). 
Whether one considers a sprawling metropolis or a small 
town, the UHI effect is linked to differences in the 
composition of the carth's surface between rural and urban 
locations. Urban development usually results in a dramatic 
alteration at the surface, as natural vegetation is removed and 
replaced by reduced evaporating, non-transpiring surfaces 
(e.g., asphalt, stone, metal, concrete). Under such alteration, 
the partitioning of incoming solar radiation into fluxes of 
sensible and latent heat is skewed in favor of increased 
sensible heat flux as evapotranspiring surfaces are reduced. 
From thermal infrared measurements (10.5-11.5 jm) 
acquired by the Advanced Very High Resolution Radiometer 
(AVHRR) aboard the NOAA series of polar orbiting satellites, 
Roth et al. (1989) derived surface temperature data and 
assessed its spatial distribution across several cities along the 
west coast of North America. Elevated daytime surface 
temperatures were highly correlated to the patterns of land 
cover related to urban land use (ie. higher surface 
temperatures corresponded to industrial areas while 
considerably cooler surface temperatures corresponded to 
well-vegetated residential areas). 
In the absence of solar illumination of the surface at night, 
however, Roth et al. (1989) found that the relationship 
between surface temperature and urban land use was ill- 
defined. Yet this is the time when the UHI effect is greatest, 
which is reflected by the finding by Karl et a/. (1988) that 
urbanization within the United States exerts its greatest 
influence on minimum (compared to. maximum or mean) 
temperature (which is mainly a nighttime phenomenon). 
Although thermal infrared satellite measurements cannot 
directly measure the UHI effect, they can be coupled with 
satellite-derived measurements of vegetation density to 
substantially describe the contributing factors to the UHI 
effect. 
Gallo et al. (1993a, 1993b) compared vegetation indices and 
radiant surface temperature acquired by the AVHRR with 
Intemational Archives of Photogrammetry and Remote Sensing. Vol. XXXII, Part 7, Budapest, 1998 
409