Determining Heat Island Response to Varying Land Cover Changes Between 2004 and 2017 Within the City of Reno, Nevada

Abstract

The objective of this research was to investigate the role of land cover changes through time in influencing spatial variability of the surface urban heat island of the metropolitan area of Reno-Sparks, Nevada. Free and widely available thermal data from Landsat 7 ETM+ (Enhanced Thematic Mapper Plus) sensor was gathered for a period between 2004 and 2017 and processed to at-satellite surface temperature. Using parcel data and the National Land Cover Database, the time series of Landsat data was sampled for areas which had undergone development during that time. This sample was cross-validated with ten iterations of equal sample size, with a mean correlation coefficient of 0.623 (standard deviation of 0.008) versus the model’s value of 0.624. A set of generalized linear models was conducted on this sample to determine expected temperature change with land cover class. It was found that recently developed regions within Reno-Sparks are 0.6 oC warmer on average than the undeveloped desert grasses and sage. When wetlands/irrigated greenery were converted to impervious surfaces, it resulted in a positive surface temperature change of over 2 oC. Once developed, no significant difference was found in the surface temperature trends. This research, using remote sensing technologies, has shown that the Reno-Sparks surface urban heat island has undergone local, but measurable growth in the last fourteen years.