Land use and land cover mapping from diverse data sources for an arid urban environments

Accurate and up-to-date data describing land use and land cover change support studies of urban growth such as quantifying the amount of rural to urban change and identifying change trajectories. This paper compares three methods for identifying urban land use/land cover, based on aerial photography, satellite imagery, and ground observations. While it might be natural to assume that classification based on ground observations would be the most accurate, this may not always be the case. Here we aim to quantify to what degree these three different classification approaches agree or contradict each other and to understand why. Land use/land cover data derived from these three sources were compared for the Phoenix metropolitan area, an arid urban region undergoing rapid urbanization. Our results show that satellite data are well suited to classify land use/land cover where land use categories are associated with homogeneous land cover at the subpixel level, but that for land use categories with subpixel land cover complexity, aerial photographs or ground observations are needed to aid in the classification.