Urban green spaces are known to provide ample benefits to human society and hence play a vital role in safeguarding the quality of life in our cities. In order to optimize the design and management of green spaces with regard to the provisioning of these ecosystem services, there is a clear need for uniform and spatially explicit datasets on the existing urban green infrastructure. Current mapping approaches, however, largely focus on large land use units (e.g., park, garden), or broad land cover classes (e.g., tree, grass), not providing sufficient thematic detail to model urban ecosystem service supply. We therefore proposed a functional urban green typology and explored the potential of both passive (2 m-hyperspectral and 0.5 m-multispectral optical imagery) and active (airborne LiDAR) remote sensing technology for mapping the proposed types using object-based image analysis and machine learning. Airborne LiDAR data was found to be the most valuable dataset overall, while fusion with hyperspectral data was essential for mapping the most detailed classes. High spectral similarities, along with adjacency and shadow effects still caused severe confusion, resulting in class-wise accuracies <50% for some detailed functional types. Further research should focus on the use of multi-temporal image analysis to fully unlock the potential of remote sensing data for detailed urban green mapping.
Spectral unmixing of urban land cover relies on representative endmember libraries. For repeated mapping of multiple cities, the use of a generic spectral library, capturing the vast spectral variability of urban areas, would constitute a more operational alternative to the tedious development of image-specific libraries prior to mapping. The size and heterogeneity of such a generic library requires an efficient pruning technique to extract site-specific spectral libraries. We propose the “Automated MUsic and spectral Separability based Endmember Selection technique” (AMUSES), which selects endmember subsets with respect to the image to be processed, while accounting for internal redundancy. Experiments on simulated hyperspectral data from Brussels (Belgium) showed that AMUSES selects more relevant endmembers compared to the conventional Iterative Endmember Selection (IES) approach. This ultimately improved mapping results (kappa increased from 0.71 to 0.83). Experiments on real HyMap data from Berlin (Germany) using a combination of libraries from different cities underlined the potential of AMUSES for handling libraries with increasing levels of generality (RMSE decreased from 0.18 to 0.15, while only using 55% of the number of spectra compared to IES). Our findings contribute to the value of generic spectral databases in the development of efficient urban mapping workflows.