见物见人：基于人群数字画像的高强度城区公园绿地使用人群精细解析

Considering both Objects and People: A Fine-grained Analysis of High-Intensity Urban Park Green Space Users Based on Digital Portraits of the Population

关键词:  风景园林  公园绿地  高强度城区  数字画像  精细解析  优化策略

Key words:  landscape architecture  park green space  high-intensity urban area  digital portrait  fine grained analysis  optimization strategy

摘要:

高强度城区是中国城镇化的典型形态，随着微气候恶化与空间压力增加，亟须公园绿地发挥游憩、调节与情绪缓释作用。但现有建设常忽视人群的实际使用规律，致使“见物不见人”的供需失衡。整合手机信令、房产交易、POI与公园空间数据，建立涵盖性别、年龄、收入3类静态属性及出行距离、停留时长、公园类型3类动态行为指标的数字画像建构框架。以南京中心城区294个公园绿地为例，采集2.18万条轨迹，绘制488类人群画像并筛选出就近休闲型、通勤游憩型等6类典型人群。结果显示，公园使用上呈高强度城区中心-外围梯度分异，社区公园在总量上显著占优，而综合公园与游园具有更高效率和包容性；时间上表现为工作日下午与周末双高峰，以短距离、短时长访问为主。地理加权回归表明，公园绿地面积、文化景点、交通与休闲POI能促进使用行为，而居住用地与就业密度则相反。基于此，提出完善高密度居住区通勤绿道、增强社区公园社交功能及外围建设特色公园等优化策略。

Abstract:

High-intensity urban development has emerged as a dominant paradigm in China's rapid urbanization process, creating concentrated centers of population and economic activity while simultaneously generating significant environmental and social challenges. These urban areas face mounting pressures, including microclimate deterioration, urban heat island effects, and heightened stress on inhabitants' psychological well-being. Urban park green spaces serve as crucial countermeasures to these challenges, providing essential ecological services, facilitating recreational activities, and contributing to mental health restoration. However, conventional planning approaches frequently prioritize physical infrastructure development over understanding nuanced human behavioral patterns, resulting in a fundamental disconnect between green space provision and actual user needs. The methodological framework incorporates diverse datasets, including anonymized mobile location data, real estate transaction records serving as socioeconomic indicators, points of interest distribution patterns, and detailed urban spatial information on park characteristics and land use configurations. The analytical process follows a five-phase approach: indicator selection based on demographic static attributes and spatiotemporal dynamic metrics, multi-source data integration with advanced geospatial techniques, indicator quantification via statistical classification, digital portrait generation through combinatorial analysis, and final extraction of prototypical user groups via dimensionality reduction. This robust analytical framework was implemented in an extensive empirical investigation of Nanjing's central urban area, encompassing 808 square kilometers of high-intensity development. The study incorporated data from 294 distinct park green spaces, analyzed over 21,800 validated visitation records via sophisticated trajectory processing algorithms, and generated 488 unique digital portraits. Subsequent statistical analysis employing principal component analysis enabled the distillation of these portraits into six predominant typologies representing distinctive usage behaviors and preferences. The research findings reveal sophisticated multi-dimensional spatial and temporal patterns of green space utilization. Analysis demonstrates a pronounced core-periphery gradient in usage intensity that closely correlates with established population density distributions. While large comprehensive parks attract the greatest absolute number of visitors due to their substantial size and diverse amenities, smaller neighborhood parks and pocket green spaces exhibit remarkably higher usage density and service efficiency, underscoring their vital role in serving high-density urban communities. Temporal analysis identifies distinct diurnal and weekly patterns with pronounced peaks during afternoon hours and weekend periods, while quantitative assessment confirms that shortdistance movements and brief duration visits constitute the dominant mode of park utilization. Advanced spatial statistical analysis through geographically weighted regression modeling identified key determinants influencing park usage patterns. The results indicate that green space area, cultural and scenic attraction density, transportation infrastructure availability, and recreational facility concentration positively correlate with increased user attendance. Conversely, residential land area extent and occupational point density demonstrate negative correlations with park usage intensity. Based on these empirical insights, the study proposes a comprehensive dual-strategy optimization framework integrating spatial configuration enhancements with targeted facility improvements. Specific recommendations include strategic implantation of pocket parks in high-density residential zones, development of quality greenways along major transportation corridors, enhancement of community parks for social interaction, and creation of specialized park facilities in peripheral urban areas. This research provides a sophisticated data-driven framework for transitioning urban green space planning from traditional supply-oriented approaches to innovative demand-responsive strategies. The digital portrait methodology facilitates detailed understanding of urban humanenvironment interactions, informs equitable, efficient and sustainable urban green space development, and effectively meets residents' needs under high-intensity developmen