Přehled o publikaci

Accurate and scalable estimation of Leaf Area Index (LAI) is essential for understanding vegetation-atmosphere interactions and assessing ecosystem services, such as air pollution removal, across spatial scales, from individual trees to entire urban landscapes. Ground-based methods provide reliable but spatially limited data, while remote sensing (RS) approaches offer broader coverage yet require extensive field validation. Neither alone sufficiently captures the structural complexity of urban vegetation. This study presents a novel framework that integrates highresolution RS data (airborne laser scanning and multispectral imagery) with hemispherical images derived from Google Street View to estimate spatially continuous LAI and model dry deposition of air pollutants. Applied in Liberec, Czech Republic, during peak vegetation season (June 2019), the framework produced LAI estimates ranging from 0.4 to 5.3 m2 m- 2 (R2 = 0.85, RMSE = 0.37) at both city-wide and local scales. The approach enabled high-resolution (1-m2) modelling of PM,0 and O3 removal across six sites representing three types of urban greenery. Urban and peri-urban forests exhibited the highest removal rates (PM,0: 2.54-2.91 g m- 2; O3: 1.85-2.07 g m- 2), followed by public green spaces (PM,0: 1.38-1.49 g m- 2; O3: 0.99-1.07 g m- 2), and street greenery (PM,0: 0.82-1.64 g m- 2; O3: 0.58-1.15 g m- 2). By integrating large- and finescale vegetation structures, this framework highlights the role of dense, complex greenery in maximizing pollution removal and demonstrates the potential of hybrid methods for precise monitoring of green infrastructure and ecosystem service assessment in urban and peri-urban settings.