UC Santa Barbara

The sensitivity of urban heat and urban vegetation to regional and global climate phenomena such as drought, heat waves, and climate change is a key concern in climate policy, urban planning, public health, and water and energy use. These signals are often difficult to resolve as urban climates vary at fine spatiotemporal scales and have complex interactions with land cover, surface morphology, and background climates. In this dissertation, I used multi-source airborne and satellite imagery from optical and thermal sensors to evaluate urban drought impacts on urban vegetation and climate and to investigate the efficacy of a novel spaceborne thermal sensor for fine-scale 24-hour monitoring of urban heat and its relationships with environmental and morphological drivers over the course of a day. In Chapter 1, I explored interactions between urban vegetation cover and urban heat during the 2012 to 2016 California megadrought in urbanized Los Angeles County, CA, USA using an 8-year time series of optical and thermal imagery. Over the course of the drought, I found strong spatiotemporally variant losses in green vegetation cover and complex coupling between losses of green cover and urban heat. In Chapter 2, I exploited the unique orbital characteristics of the NASA ECOsystem Spaceborne Thermal Radiometer Experiment on Space Station (ECOSTRESS) sensor to develop composited 24-hour urban land surface temperature (LST) imagery for diurnal analysis of urban heat. I found that ECOSTRESS composites were suitable for resolving fine-scale (in both space and time) interactions between environmental drivers and urban LST over a full diurnal cycle as well as higher order features such as variability in heating and cooling rates based on land cover type. Finally, in Chapter 3, I used fine-scale urban surface morphology data from LiDAR with a parameterization of subpixel sun-surface-sensor geometry to extract facet-scale urban LSTs (e.g., wall, roof, road) from a 2.5-year time series of ECOSTRESS imagery. I then tracked the diurnal course of thermal anisotropy in urbanized Los Angeles County and New York City, NY, USA finding strong spatiotemporal variability in angular effects on urban LST as a function of surface-sensor geometry and surface morphology as well as separation in facet-scale LST based on facet orientation and time of day. These findings highlight the sensitivity of urban vegetation and climate to drought and potential tradeoffs between efforts to increase urban green cover and water conservation needs under a warming and drying climate. In addition, this work suggests that satellite platforms with atypical sampling regimes (e.g., from a precessing orbit) can provide data that is unique and appropriate for characterizing otherwise undersampled dimensions of urban thermal climates.