科学高效地配置和调度消防救援资源是公共安全领域的重大需求。现有研究较少考虑实时路况对救援车辆行驶路线选择及行驶时间的影响, 且忽视了消防救援可覆盖范围的非连续分布特征。该研究发现了消防救援可覆盖范围非连续分布而导致的“救援飞地”现象, 并将其定义为“不与消防站直接毗连但依然在该消防站有效救援范围内的片区”, 进而开发了基于实时路况的精准识别救援飞地及其面积的算法。为验证该算法的可行性, 以CS-XX城市消防站为例, 抓取了3 818个兴趣点作为救援需求点, 并设置了49个评估时刻, 获取了187 082条有效数据, 进而计算出CS-XX消防站的实际可覆盖面积和救援飞地面积。结果表明该文提出的算法可有效识别救援飞地：CS-XX消防站的可覆盖面积受交通路况影响, 范围在1.83~4.57 km²之间; 该站可通过临近的过河桥梁覆盖对岸的部分区域, 形成了救援飞地; 救援飞地位置超出了传统面积确定法的7.00 km²圆形区域范围, 且其面积最大时可达到CS-XX消防站可覆盖面积的27.53%。建议将救援飞地作为消防救援责任辖区, 进而提高既有消防救援资源的利用效率。

Objective: The efficient allocation and dispatch of fire rescue resources are crucial to urban public safety. Traditional approaches assume continuous spatial distribution of fire service coverage areas and give less consideration to the impact of real-time traffic conditions on rescue route selection and response times. This study aims to introduce and define the concept of "rescue enclaves"—areas that, although not directly adjacent to fire stations, can be effectively covered by them—and proposes a method to identify and calculate these spatially discontinuous coverage areas. Methods: This study proposed a method for identifying and calculating spatially discontinuous coverage areas by mapping points to grids. Using this method: (1) fire truck travel times were calculated using real-time traffic data, (2) geographic coordinates were converted to universal transverse Mercator (UTM) coordinates, (3) the region was divided into fine grids, (4) grid coverage status was determined, (5) transition grids were processed through neighborhood analysis, and (6) rescue enclaves were identified using a breadth-first search (BFS) algorithm. The CS-XX urban fire station in a Chinese city was selected as a case study to validate the method. In this case study, 3 818 points of interest were identified as rescue demand points across 49 evaluation periods in one day, generating 187 082 valid data samples. A target response time of 4 min was established, and an 80% reduction coefficient was applied to convert regular vehicle travel times to fire truck travel times. Results: The rescue enclave areas were successfully identified and calculated using the proposed method, through which the following key findings were revealed: (1) the dynamic coverage area of CS-XX was observed to vary from 1.83 to 4.57 km², with the minimum fire service coverage of 1.83 km² being recorded during the morning peak at 8:00, (2) the calculated coverage area trends were found to be consistent with the percentage of demand points accessible within 4 min, whereby the reliability of the method was validated, (3) critical rescue enclaves were identified near CS-XX, with enclave areas ranging from 0.25 to 1.12 km², accounting for 12.20%-27.53% of the total coverage area, (4) the rescue enclaves were observed to occasionally extend beyond the traditional coverage of 7.00 km² prescribed by standard area determination methods, and (5) coverage areas and rescue enclave areas were demonstrated to synchronously vary with traffic conditions, with traffic congestion leading to a significant reduction in their sizes. Conclusions: The proposed conceptualization of rescue enclaves is elucidated in this study, and their substantial manifestation within fire service coverage areas is substantiated through rigorous analysis. The rescue enclaves are systematically identified and quantified via an algorithmically driven methodological framework, and it is ascertained that such enclaves may comprise up to 27.53% of the coverage area of a fire station. If rescue enclaves are integrated into fire rescue jurisdiction planning protocols, they can substantially optimize resource allocation efficacy. While real-time traffic conditions and different flow efficiencies across heterogeneous route typologies are identified as the primary determinants of enclave formation, subsequent investigations are warranted to elucidate the precise mechanistic underpinnings and contributory factors governing rescue enclave emergence as well as to establish quantitative metrics for rescue passage efficiency across diverse route configurations.