Urban green spaces (UGS) are crucial urban elements that serve as direct carbon sequestration and contribute to indirect carbon emission reduction. Accurately calculating the carbon density of urban green spaces allows for scientific planning and design, thereby advancing efforts toward achieving carbon neutrality. This study has developed a workflow for estimating carbon density in urban green spaces through point cloud measurements and model simulations, using the UGS in the Beijing Main District as a case study. From the sample level, a calculation methodology was constructed based on the point cloud technology-model simulation method, which can obtain the carbon density at the plant level and the sample level. At the UGS level, remote sensing inversion was utilized to map the carbon density of urban green spaces. Ultimately, the research calculated and compared carbon density at different scales, including the carbon density of individual plants, the carbon density of sample plots, and the carbon density of various types of urban green spaces. It was found that the carbon density of trees in UGS was 9.87 kg/m², while those of shrubs and herbaceous plants were 13.20 kg/m² and 0.11 kg/m². In urban green spaces, the carbon densities of the tree and herb layers were slightly lower than those in natural ecosystems, whereas the carbon density of the shrub layer was significantly higher. This highlights the substantial potential and value of shrubs in carbon sequestration and carbon storage. The average carbon density of all UGS types was 9.76 kg/m², with the following descending order: Neighborhood Parks (10.31 kg/m²) > Attached Green Spaces (7.22 kg/m²) > Regional Parks (5.75 kg/m²). Based on these findings, the study proposed optimization strategies for different UGS types, focusing on high carbon-density plant community optimization. The goal is to provide a theoretical foundation for carbon storage calculations and plant arrangements in future UGS construction.

As part of urban green space and emergency shelters, refuge green spaces (RGS) contribute significantly to the resilience of cities to natural disasters. In contrast, few studies have been conducted to assess the equity of RGS in relation to their planning layout. The presented research aims to quantitatively evaluate the equity of RGS within Chengdu’s Third Ring Road, and to propose corresponding optimization measures in conjunction with future green space planning. The rapid evacuation capacity of the RGS was evaluated by calculating the equity of the RGS in walking modes of 5, 10 and 15 min using an improved three-step floating catchment area method (3SFCA). Based on the results, RGS had an average equity in the study area. The total number of RGS within the Third Ring Road of Chengdu was insufficient, with an uneven spatial distribution and a structure to be optimized. The rapid evacuation capacity of RGS in 5 and 10 min needs to be further improved. The short-term resettlement capacity of RGS after 15 min was relatively good. Using Moran’s I index, the RGS equity and house price results were analyzed and no significant aggregation and polarization were observed. Following the evaluation, recommendations are made for optimizing and adding future RGS in accordance with Chengdu Green Space System Planning. The equity of RGS has been significantly improved after optimization, which can meet the needs of over 90% of residents for 15 min. This study provided feasible suggestions for the layout and structural optimization of the future RGS within the Third Ring Road of Chengdu, which aimed to create a RGS network with complex functions, to meet the multiple needs of citizens.

The visual attributes of urban green spaces influence people’s perceptions, preferences, and behavioural activities. While many studies have established correlations between landscape perception and visual attributes, they often focus on specific timeframes and overlook dynamic changes in the spatial form of urban green spaces. This study aims to explore the long-term changes in the visual attributes of urban green spaces. We propose a method to quantitatively analyse changes in visual attributes using point clouds to simulate visual interfaces. Using an unmanned aerial vehicle, we conducted a five-axis tilt photography survey of Qinglvyuan Park in Nanjing, China, in August 2018 and September 2023. Point cloud models were generated for the two periods, and five visual attribute indicators, openness (OP), depth variance (DV), green view ratio (GVR), sky view ratio (SVR), and skyline complexity (SC), were analysed for long-term changes. The results indicate that OP, DV, and SVR decreased after five years, while GVR increased. The maximum increase in GVR was 26.6%, and the maximum decrease in OP was 12.8%. There is a positive correlation between GVR and its change (d_GVR). Conversely, there are negative correlations between SC and its change (d_SC), as well as between SVR and d_GVR. Tree growth emerged as a primary factor influencing changes in the visual attributes of urban green spaces. This study highlights the importance of adopting a long-term and dynamic perspective in visual landscape studies, as well as in landscape design and maintenance practices. Future research on predicting long-term changes in the visual attributes of urban green spaces should focus on understanding the relationships between tree properties and environmental conditions.

Urban green spaces are pivotal to enhancing carbon sinks and advancing carbon neutrality goals, yet the structural complexity of green space units often leads to scale mismatches and weak spatial responsiveness in current assessments. This study develops an integrated evaluation framework that combines landscape spatial unit typologies with life-cycle-based carbon flux modeling. We defined 22 landscape spatial unit types based on two-dimensional surface cover and three-dimensional vegetation structure, including waterbodies and vertical greening. A life-cycle carbon model was developed with indicators covering unit carbon sink, unit carbon emission, unit net carbon sink efficiency, and time to carbon balance. Taking Luhe Park in Nanjing as a case study, the carbon sink efficiency indicators were quantified for 108 units over a 50-year cycle. Results indicate that multilayer vegetation structures, high green coverage, and moderate-to-high planting density markedly enhance carbon sink efficiency, whereas extensive built surfaces and high impervious ratios suppress it. K-means clustering classified the spatial units into four types with emphasis on efficiency-driven, structural optimization, functional compatibility, and imbalance compensation, respectively, revealing a clear gradient tied to spatial configuration. To translate diagnosis into design, we report 95% confidence intervals of key structural factors as actionable thresholds. These ranges inform targeted interventions such as maintaining continuity and multilayer structure in high-efficiency areas, modest structural upgrades with native drought-tolerant plants, edge greening with permeable pavements in open spaces, and streamlined vertical systems linked to adjacent high-sink ground units. The framework delivers spatially explicit, life-cycle-aware evidence to support low-carbon planning and design of urban green spaces.

Accurately estimating aboveground carbon storage (AGC) of urban vegetation remains a major challenge, due to the heterogeneity and vertical complexity of urban environments, where traditional forest-based remote sensing models often perform poorly. This study integrates multimodal remote sensing data and incorporates two three-dimensional structural features—mean tree height ($H_{mean}$) and canopy cover ratio (CCR)—in addition to conventional spectral and textural variables. To minimize redundancy, the Boruta algorithm was applied for feature selection, and four machine learning models (SVR, RF, XGBoost, and CatBoost) were evaluated. Results demonstrate that under multimodal data fusion, three-dimensional features emerge as the dominant predictors, with XGBoost using Boruta-selected variables achieving the highest accuracy (R² = 0.701, RMSE = 0.894 tC/400 m²). Spatial mapping of AGC revealed a “high-aggregation, low-dispersion” pattern, with the model performing best in large, continuous green spaces, while accuracy declined in fragmented or small-scale vegetation patches. Overall, this study highlights the potential of machine learning with multi-source variable inputs for fine-scale urban AGC estimation, emphasizes the importance of three-dimensional vegetation indicators, and provides practical insights for urban carbon assessment and green infrastructure planning.