Karl, T. R., H. F. Diaz, and G. Kukla, 1988: Urbanization: Its detection and effect in the United States climate record. J. Climate, 1, 1099-1123.
Several equations were developed that related the effect of urban growth, measured by increasing population, to the mean seasonal and annual temperature: the diurnal maximum, minimum, average, and range. These equations were derived from a network of 1219 stations across the United States, which were analyzed for the years 1901-84. The results indicate that urban effects on temperature are detectable even for small towns with populations under 10,000. Stations with populations near 10,000 are shown to average 0.1°C warmer for the mean annual temperature than nearby stations located in rural areas with populations less than 2000. Urbanization decreases the daily maxima in all seasons except winter and the temperature range in all seasons. It increases the diurnal minima and the daily means in all seasons.
The equations indicate that, for the annual mean temperature, urbanization during the twentieth century accounts for a warm bias of about 0.06°C in the U.S. Historical Climatology Network (HCN). Due to the large number of stations located in sparsely populated arms [(over 85% (70%) of all stations had a 1980 population of less than 25,000 (10,000)], the impact of urbanization is not large in relation to decadal changes of temperature in the United States. The average heat island impact during the period 1901-84 for the HCN is largest for the daily minima (0.13°C) and the temperature range (0.14°C), while the impact on the daily maxima (0.01°C) is an order of magnitude smaller.