Hyperspectral satellite imagery for urban climate applications

Urban climate research relies on multispectral (MS) satellite imagery because of its global coverage and relatively high spatial and temporal resolution. However, its coarse spectral detail limits the analysis of complex urban surfaces. New hyperspectral (HS) satellite missions provide much finer spectral information, supporting detailed analysis of urban microclimates. Here, we present an overview of current HS satellite products and examine the potential of PRISMA (PRecursore IperSpettrale della Missione Applicativa) and DESIS (German Aerospace Center (DLR) Earth Sensing Imaging Spectrometer) missions for two key applications in urban environments: material abundance estimation and local climate zone (LCZ) mapping. To estimate material abundances, we apply constrained spectral unmixing to HS imagery over Milan, Italy. Results are compared with near-simultaneous MS data and validated against the local geotopographic database. The derived abundances are also linked to high-resolution air temperature maps predicted using a machine learning-based regression approach. Secondly, LCZs are mapped using a combined RS and GIS-based method, integrating spectral and spatial information for improved classification of urban areas.

Our results show that HS imagery supports sub-pixel material estimation, opening the possibility for a transition from single land-cover labels to multi-material representations within each pixel. Thermal assessment further validated these estimates, with natural materials reducing heat and artificial surfaces increasing it. Finally, LCZ mapping resulted in higher accuracy with HS imagery compared to MS products.

HS imagery provides a promising path for applications in urban climate research and other urban studies. Thanks to its technical advantages over MS imagery, HS data enable the generation of data suitable for microclimate modelling, heat mitigation assessment or urban management. Although HS imagery is not yet as widely available as MS, upcoming missions are steadily expanding access for scientific use in urban monitoring.