Electronic Thesis and Dissertation Repository

Thesis Format



Master of Science




Voogt, James A.


Urban surface temperatures are important variables in urban climatological processes. This thesis examines the directional variability of remotely sensed urban surface temperatures (thermal anisotropy or Λ) for three vegetated residential neighbourhoods in Salt Lake City, Utah, USA. Airborne thermal remote sensing using a thermal imager sampled the directional brightness temperature (DBT) at three times within a day for each site. Results indicate that temporal variability over a 20 – 30-minute flight was not negligible. Average DBT were then extracted from atmospherically corrected images and plotted on polar plots. For low density residential neighbourhoods Λ is increased with increasing tree-canopy coverage (λtree) due to the increased temperature contrast. The ΛMax for the sites with large λtree were ~8°C compared to ~6°C for the site with sparse λtree. These results indicate Λ for low density residential neighbourhoods is significant and must be considered when discussing land surface temperatures for similar sites.

Summary for Lay Audience

Land surface temperatures are an important variable for many climatological and hydrological processes. The urban surface, however, is a rough 3-dimensional surface that is comprised of many different surface materials that can be sunlit or shaded. A remote sensing instrument observing this surface can only observe a subset of these surfaces, and as the sensor is moved around the site, it will continue to see a different combination of surfaces and consequently report different averaged surface temperatures. This variability in observed directional surface temperatures is the thermal anisotropy. This thesis examines the thermal anisotropy for residential neighbourhoods in three typical North American neighbourhoods with varying amounts of tree canopy coverage located in Salt Lake City, UT, USA. To sample the thermal anisotropy, airborne thermal remote sensing using a thermal imager was used. Flights were conducted at three times a day for each site. Thermal imagery were corrected for atmospheric effects using MODTRAN6 and atmospheric profile data compiled from a microwave radiometer and radiosondes launched as part of this project. Average temperatures were calculated for each view direction and angle combination for each flight and plotted to illustrate the directional variability of the observed average surface temperature. The temporal variability of surface temperature over a 20 – 30-minute flight was not negligible, and this thesis presents a method of correcting for this variability to minimize impacts on the derived anisotropy. Results indicate that the maximum daytime thermal anisotropy (Λ) of these neighbourhoods’ ranges between 6 – 8°C, with the smallest anisotropy observed for the site with lowest tree-canopy coverage. The other two sites had similar tree-canopy coverage and slight differences in building coverage but very similar Λ. This indicates that large amounts of tree canopy coverage create larger thermal anisotropy in low density residential neighbourhoods. It was found that this is likely due to the randomly located cooling effect of the tree canopies causing cooler and larger shaded regions in contrast with the hot surfaces such as sunlit roofs and roads. The large anisotropy of these spatially extensive neighbourhoods is potentially important to the application of satellite-based thermal remote sensing of urban areas.

Creative Commons License

Creative Commons Attribution 4.0 License
This work is licensed under a Creative Commons Attribution 4.0 License.