【目的】随志愿者地理信息系统的快速发展,高现势性众源路网已成为智慧城市建设的重要数据来源,其选取的效率与效果成为影响多尺度数据服务的关键因素。已有的路网选取方法大多基于数据属性信息判断道路重要性,十分合理且有效,但是,实...【目的】随志愿者地理信息系统的快速发展,高现势性众源路网已成为智慧城市建设的重要数据来源,其选取的效率与效果成为影响多尺度数据服务的关键因素。已有的路网选取方法大多基于数据属性信息判断道路重要性,十分合理且有效,但是,实际数据往往存在属性缺失问题,一定程度上限制了方法的适用性。【方法】针对此问题,本文提出一种属性信息缺失条件下的众源路网空间句法自动建模与选取方法。首先,基于开放街道地图(Open Street Map)中心线数据,开发程序自动执行几何化简、拓扑修正与伪节点处理,批量生成整个城市的空间句法线段模型,并基于模型计算整合度、选择度等空间句法指标;随后构建Stroke,并提取几何特征;进一步,创新性地提出2项复合指标:基于路径单元的标准化角度整合度(SNAIN)与基于路径单元的标准化角度选择度(SNACH),以联合刻画道路的拓扑可达性与几何连续性。在此基础上,应用结合熵权法与层次分析法(EW-AHP)的主客观集成赋权方法,确定综合指标的权重,实现道路的重要性排序。最后,通过断头路识别与网格密度修补,进一步提高路网的连通性和完整性。【结果】以兰州(带状道路网)和成都(环形放射状道路网)为案例验证,结果表明:在道路属性信息缺失的条件下,本文方法能够有效识别城市主干路网,其与OSM道路等级匹配准确率分别达到兰州0.9421、成都0.9711;修补后兰州市路网连通率由1.0582提升至1.0864,成都市路网连通率由1.1086提升至1.1198(成都在所选尺度内的断头路完全消除)。消融实验表明,SNAIN更有利于提升全局连通性,SNACH有助于增强几何连续性,二者并用能在连通性与空间覆盖间取得平衡。【结论】本文方法为属性信息不完整情形下的大规模城市路网快速建模与选取提供了新的理论支持和技术路径。展开更多
Biological collections are critical for the understanding of species distributions and for formulating biodiversity conservation strategies.However,biological collections are susceptible to various biases,including th...Biological collections are critical for the understanding of species distributions and for formulating biodiversity conservation strategies.However,biological collections are susceptible to various biases,including the“road-map effect”,meaning that the geography of biological collections can be influenced by road networks.Here,using species occurrence records derived from 921,233 plant specimens,we quantified the intensity of the“road-map effect”on floristic collections of China,and investigated its relationships with various environmental and socio-economic variables.Species occurrence records mainly distributed in major mountain ranges,while lowlands were underrepresented.The distance of species occurrence records to the nearest road decreased from 19.54 km in 1960s to 3.58 km in 2010s.These records showed significant clustering within 5 km and 10 km buffer zones of roads.The road density surrounding these records was significantly higher than that in random patterns.Collectively,our results confirmed a significant“road-map effect”in the floristic collections of China,and this effect has substantially intensified from the 1960s to the 2010s,even after controlling for the impact of road network expansion.Topographic,climatic and socio-economic variables that determine regional species diversity,vegetation cover and human impact on vegetation played crucial roles in predicting the intensity of the“road-map effect”.Our findings indicate that biological surveys have become increasingly dependent on road networks,a trend rarely reported in published studies.Future floristic surveys in China should prioritize the lowland areas that have experienced stronger human disturbances,as well as remote areas that may harbor more unique and rare species.展开更多
文摘【目的】随志愿者地理信息系统的快速发展,高现势性众源路网已成为智慧城市建设的重要数据来源,其选取的效率与效果成为影响多尺度数据服务的关键因素。已有的路网选取方法大多基于数据属性信息判断道路重要性,十分合理且有效,但是,实际数据往往存在属性缺失问题,一定程度上限制了方法的适用性。【方法】针对此问题,本文提出一种属性信息缺失条件下的众源路网空间句法自动建模与选取方法。首先,基于开放街道地图(Open Street Map)中心线数据,开发程序自动执行几何化简、拓扑修正与伪节点处理,批量生成整个城市的空间句法线段模型,并基于模型计算整合度、选择度等空间句法指标;随后构建Stroke,并提取几何特征;进一步,创新性地提出2项复合指标:基于路径单元的标准化角度整合度(SNAIN)与基于路径单元的标准化角度选择度(SNACH),以联合刻画道路的拓扑可达性与几何连续性。在此基础上,应用结合熵权法与层次分析法(EW-AHP)的主客观集成赋权方法,确定综合指标的权重,实现道路的重要性排序。最后,通过断头路识别与网格密度修补,进一步提高路网的连通性和完整性。【结果】以兰州(带状道路网)和成都(环形放射状道路网)为案例验证,结果表明:在道路属性信息缺失的条件下,本文方法能够有效识别城市主干路网,其与OSM道路等级匹配准确率分别达到兰州0.9421、成都0.9711;修补后兰州市路网连通率由1.0582提升至1.0864,成都市路网连通率由1.1086提升至1.1198(成都在所选尺度内的断头路完全消除)。消融实验表明,SNAIN更有利于提升全局连通性,SNACH有助于增强几何连续性,二者并用能在连通性与空间覆盖间取得平衡。【结论】本文方法为属性信息不完整情形下的大规模城市路网快速建模与选取提供了新的理论支持和技术路径。
基金funded by National Natural Science Foundation of China(32460276,32060275)Jiangxi Provincial Natural Science Foundation(20232BAB203058,20242BAB27001)。
文摘Biological collections are critical for the understanding of species distributions and for formulating biodiversity conservation strategies.However,biological collections are susceptible to various biases,including the“road-map effect”,meaning that the geography of biological collections can be influenced by road networks.Here,using species occurrence records derived from 921,233 plant specimens,we quantified the intensity of the“road-map effect”on floristic collections of China,and investigated its relationships with various environmental and socio-economic variables.Species occurrence records mainly distributed in major mountain ranges,while lowlands were underrepresented.The distance of species occurrence records to the nearest road decreased from 19.54 km in 1960s to 3.58 km in 2010s.These records showed significant clustering within 5 km and 10 km buffer zones of roads.The road density surrounding these records was significantly higher than that in random patterns.Collectively,our results confirmed a significant“road-map effect”in the floristic collections of China,and this effect has substantially intensified from the 1960s to the 2010s,even after controlling for the impact of road network expansion.Topographic,climatic and socio-economic variables that determine regional species diversity,vegetation cover and human impact on vegetation played crucial roles in predicting the intensity of the“road-map effect”.Our findings indicate that biological surveys have become increasingly dependent on road networks,a trend rarely reported in published studies.Future floristic surveys in China should prioritize the lowland areas that have experienced stronger human disturbances,as well as remote areas that may harbor more unique and rare species.
