特大城市
绝对偏差
环境科学
城市热岛
标准差
热的
熵(时间箭头)
对比度(视觉)
大气科学
数学
平均绝对误差
统计
暴露持续时间
强度(物理)
气象学
大偏差理论
热容
条件熵
显热
平均差
暴露评估
计量经济学
地理
街道峡谷
作者
Tao Wu,Zeyin Chen,Shujie Yang,Runyao Huang,Ruhang Wei,Jie Cao,Siying Li,Zhiqiang Wu
标识
DOI:10.1016/j.eiar.2026.108802
摘要
Urban heat exposure in highly mobile cities cannot be fully characterized by residential conditions alone. Existing mobility-based assessments quantify absolute exposure or residential-to-mobile differences, but do not separate population-level exposure averaging from individual departures around that relationship. We develop a three-layer framework distinguishing absolute exposure, time-matched out-of-home thermal contrast, and conditional deviation from the Neighborhood Effect Averaging Problem (NEAP) relationship. NEAP-adjusted heat exposure deviation (NAE) measures the difference between observed mobile heat exposure (MHE) and the MHE expected given residential heat exposure (RHE). Using reconstructed hourly land surface temperature and mobile-phone trajectories for 179,588 commuters in Shanghai, we identify clear population-level averaging between RHE and MHE ( β = 0.730 ). NAE and the direct out-of-home contrast differed in direction for 32.3% of commuters. Elevated NAE formed a distinct peri-core and commuting-transition geography rather than reproducing absolute heat hotspots; 26.8% of upper-decile NAE grids were not simultaneously identified by high RHE, MHE, or direct out-of-home contrast. Higher NAE was associated with longer home–work paths, longer commuting duration, and more time spent away from home, whereas greater mobility entropy was associated with lower NAE. Higher-rent residential contexts and grids with mean commuter ages in the early-to-mid 30s also showed elevated NAE, while greater functional richness and development intensity were associated with lower NAE across parts of their observed ranges. These findings establish NEAP as an empirical reference for revealing structured mobility-related exposure deviations that remain hidden in conventional absolute or direct-contrast assessments.
科研通智能强力驱动
Strongly Powered by AbleSci AI