Association between land use and temperatures during the 2011 heat wave in Houston: A quantile regression analysis
Quantile regression is an analysis tool that can investigate the effects of the covariates on a range of conditional quantiles of the responses, and thus offer a more complete view of the association between covariates and responses than the traditional linear regression. In this thesis project, we adopted the quantile regression approach to investigate the association between land cover and hot temperature by analyzing the 2011 Houston heat wave data. We estimated the effects of several classes of land cover variables on the heat wave, and conducted statistical tests to assess the constancy and overall significance of these effects across quantiles of the hot temperature. Our results show that, in the metropolitan Houston area, a location with longer distance to the Gulf of Mexico and larger percentage of developed high intensity surface may have higher temperatures during the heat wave than other locations. In addition, the difference in the temperatures between those "hotter" locations and others may be even amplified during the hottest days.
Zhou, Weihe, "Association between land use and temperatures during the 2011 heat wave in Houston: A quantile regression analysis" (2013). Texas Medical Center Dissertations (via ProQuest). AAI1549853.