A systematic approach to model the influence of the type and density of vegetation cover on urban heat using remote sensing

Cities around the world are pursuing increasing green or vegetation cover as a way of managing heat whilst improving beauty, biodiversity and recreational value. However, the pattern of the relationship between vegetation cover and urban temperature can be masked, controlled or exaggerated by vegetation structure, topography and other climate variables. This study examines the relationship between Sydney's urban surface temperature and vegetation cover as defined by two vegetation indices; mixed vegetation cover and tree cover exclusively. The shape of this relationship and relative influence of confounding factors are explored using penalised-likelihood criteria ranked regressions. Overall, increasing tree cover reduces average surface temperatures more dramatically than mixed vegetation cover. This study demonstrates that the extent of influence of greencover on surface temperatures is more accurately defined by identifying and incorporating site specific factors that confound the influence. Best predictor models are significantly improved when the influences of elevation, coastal effects and urban structure are added. Therefore, heat reducing urban greening strategies will be improved if based on local variables and conditions.