A Simplified Framework for High-Resolution Urban Vegetation Classification with Optical Imagery in the Los Angeles Megacity

Excessive fossil fuel CO2 emissions and high ozone levels in Los Angeles, California make the Southern California Air Basin (SoCAB) one of the most heavily polluted regions in the United States. However, the impact of the biosphere on CO2 levels in Los Angeles is a major source of uncertainty in estimating anthropogenic emissions. Compared to other megacities, a relatively large fraction of SoCAB consists of vegetated area, and the biospheric CO2 flux in Los Angeles is non-negligible, potentially making up 15% of the excess flux at certain times of the year. To improve estimates of biospheric contributions to the urban carbon budget, high spatial resolution maps are needed to capture the heterogeneity of urban land cover while spanning the regional domain used in carbon inversion models. We present a method of data fusion using supervised classification for developing a high spatial resolution map of urban vegetation cover in SoCAB with publicly available satellite imagery. This technique uses Sentinel-2 (10-60 m x 10-60m) and NAIP (60 cm x 60 cm) imagery to classify urban and non-urban areas of impervious surface, tree, grass, shrub, non-photosynthetic vegetation, pools, and lakes. Our approach was developed for Los Angeles, a geographically complex megacity characterized by diverse Mediterranean land cover and a mix of high-rise buildings and topographic features that produce strong shadow effects. We show that a fused NAIP and Sentinel-2 classification reduces misclassified shadow pixels and resolves spatially heterogeneous vegetation gradients across urban and non-urban regions in SoCAB at 60 cm resolution with 83% accuracy. Our approach thus provides flexibility and accuracy for application in any urban region. Results from this study will enable long-term monitoring of land cover change associated with urbanization, and quantification of biospheric contributions to carbon and water cycling in cities.