4.6.6: Land Cover and Racial Segregation

In the United States extreme heat is responsible for about 20% of natural hazard deaths (Borden and Cutter 2008). Some studies have concluded that there are large racial disparities in heat-related deaths while other studies have not found any difference (Greenberg et al. 1983; Jones et al. 1982; Kaiser et al. 2007; O’Neill et al. 2005; Schwartz 2005; Ramlow and Kuller 1990; Weisskopf et al. 2002; Ellis et al. 1975). The lack of vegetation, particularly trees, in urban areas may contribute to these disparities (Uejio et al. 2011; Hart and Sailor 2009; Oke 1982, 1989). Urban trees are known to be helpful for preserving a healthy urban environment. Some of the benefits include providing shade on hot days, reducing the need for wastewater treatment, reduced air pollution, and reduced noise pollution. In fact, urban tree coverage has been associated with lower overall mortality rates and positive health outcomes (Givoni 1991; Scott et al. 1999; Keim et al. 2006; Hwang et al. 2011; Nowak and Greenfield 2010; Samara and Tsitsoni 2011; Mitchell et al. 2011; Dadvand et al. 2012).

In the United States many studies have found racial and ethnic disparities in the presence of urban trees (Heynen 2006; Landry and Chakraborty 2009; Lowry et al. 2012; Ogneva-Himmelberger et al. 2009; Perkins et al. 2004; Zhang et al. 2008). These studies have tended to focus on single metropolitan areas. A national study from 2013 provided the first nationwide evidence of these trends (Jesdale et al. 2013). The presence of trees and other types of land cover were obtained from the 2001 National Land Cover Dataset (Homer et al. 2004). This data contains information at the census block level. According to the U.S. Census Bureau, census blocks are areas bounded by visible features such as roads, streams, and railroad tracks, and by non-visible boundaries such as property lines and city and county limits. Population data, household income, and home ownership were obtained from the 2000 U.S. census.

One interesting issue in this study is that the presence of tree cover in a census block may be uneven, and a resident of a census block may not live in the area where the trees are located. This effect may be especially true in rural areas, and for that reason the researchers only considered residents of metropolitan areas with census block population densities of more than 2,000 people residing within one square kilometer. The analysis was further restricted by the available data. That is, an area could be used if both tree cover information, demographic, and economic data were available. Therefore, the researchers restricted their analysis to residents for whom block-group level poverty information was available, and who identified as either Hispanic, non-Hispanic white, African American, or Asian (Jesdale et al. 2013).

The question of what population is being studied here is quite interesting. From the onset the basis of the study is a meeting of two major sources of data: the 2001 National Land Cover Dataset and the 2000 U.S. Census. Both databases contain a large amount of information. The National Land Cover Dataset contains land coverage information over the entirety of the United States using data based on information obtained by satellite imaging. The U.S. Census also encompasses the entire United States, so the linking of this data would potentially include the entire population, though the data is aggregated by census blocks. Because of the technical issues described above, the researchers then reduced the scope of the study to densely populated urban areas. The second restriction came from the fact that while the census is meant to encompass the entire United States, extended data is not collected in all areas; hence only the areas that collected the demographic and economic information that was relevant to the study were used. Therefore, from one viewpoint the population is restricted only to densely populated urban areas that were studied by the U.S. Census for extended demographic and economic data. This is a difficult population to define, but there is some help. As will be discussed later in the book, there is a way to choose the census blocks in a way so that they represent what would occur in the entire population of the United States. The methods used by the Census Bureau are selected to ensure this, so we can conclude that the population studied for this research corresponds to densely populated urban areas in the United States.