Community-scale microclimate simulation using Airborne Laser Scanning and object-based urban tree classification

Urban climate and urban heat island effects, resulting from urban development and expansion, significantly impact human health and well-being. Urban forests play a vital role in regulating urban climate by cooling the ambient environment through shade and evapotranspiration, among many essential ecosystem services. Different human communities have distinct abundances, compositions, and patterns of urban forests, leading to varied cooling effectiveness and microclimate outcomes. While the general consensus is that planting more trees yields greater benefits, there is an increasing need to understand and document the forests’ varied taxonomical, structural, and biophysical properties at a more granular scale. This understanding is crucial to predicting and comparing the consequent microclimate conditions across communities. In this study, we utilize high-density airborne point cloud data (58 pulses per m²) to differentiate trees based on their vertical and internal structures. We create a tree family and size classification map for all trees in two socioeconomically distinct communities in Portland, Oregon, USA. The Random Forest classifier, using Lidar-derived metrics, classifies all tree objects into seven classes with an overall accuracy of 67.1 %. Using the classified properties, we simulate and compare these forests’ combined air temperature cooling effects and test alternative tree composition scenarios with the object-based ENVI-met model. The simulation results indicate that wealthier communities experience a more significant reduction in 1.5-meter-above-ground air temperature than less wealthy communities (0.23 K cooler on average). This disparity in benefits is likely to widen further if current forest evolution trends persist. This study demonstrates a comprehensive workflow from generating object-based knowledge on urban trees to high spatial resolution microclimate simulation of the urban forests’ composite effects. This approach can aid in practical urban forest quality monitoring and evaluation, supplement urban planning and management practices, and optimize the urban forest to facilitate cross-community ecosystem services equity.