空间微观模拟方法及在城市研究中的应用

基于活动主体的城市系统微观模拟可能在未来城市研究中发挥重要作用,但其通常受到微观个体数据稀缺的限制。空间微观模拟方法（spatial microsimulation）主要基于家庭、个人等微观分析单元,通过整合不同层面的数据源,如宏观汇总层面的人口普查统计表以及微观层面的家庭活动日志调查等,合成大样本微观个体数据集,可以在精细化空间尺度上对微观个体行为进行模拟研究。该方法在城市系统微观模拟、空间分析以及政策评估等方面具有一定优势,在西方国家城市研究中的应用逐渐增多,但在国内较为缺乏。本文尝试对空间微观模拟方法的起源、三种核心算法,包括条件概率（conditional probability） 、确定性加权（deterministic reweighting）以及模拟退火（simulated annealing）进行介绍,并从国际层面综述该方法在城市研究,如收入与贫困、交通出行、健康等领域中的应用,为我国相关研究的开展提供借鉴。     

职住分离的空间差异性及其影响因素研究

制度转型与空间重构背景下,中国大城市的居住与就业空间关系发生明显变化,职住分离 现象逐渐凸显。尽管城市地理学者逐步关注转型期城市居住与就业空间关系的变化对居民通 勤行为的影响,然而深入探讨微观个体所承受的职住分离程度的差异性及其影响因素的实证研 究仍然缺乏。基于对北京城市520 户家庭、806 个通勤样本的问卷调查数据,本文借助多元回归 模型验证了居住区类型、家庭及住房状况、以及其他社会经济属性等居民职住分离程度差异性 的影响,从而折射城市转型过程中宏观制度性及结构性因素(如土地市场化改革、住房政策、单 位制度改革、城市空间扩展等) 对个体日常生活经历差异性的影响。     

基于多智能体模型与建筑物信息的高空间分辨率人口分布模拟

自上而下的人口分布模拟模型自动化程度较低,难以分析人口分布成因,且因精细尺度的人口样本较难获取而不太适用于高空间分辨率人口分布模拟。提出了一种基于多智能体模型和建筑物信息的高空间分辨率人口分布模拟模型。首先利用建筑物三维分布数据提取住宅建筑,构建智能体人口分布模拟模型的环境;然后基于统计、调研数据定义智能体属性,确定智能体居住选择行为规则;最后以泰日社区为例进行了居住人口分布仿真模拟。研究结果表明,基于建筑物信息的人口分布多智能体模型,可以获取每栋建筑物上的人口,改进了当前高分辨率人口模拟主要只模拟小区或者居委会人口的不足;多智能体模型具有较高的自动化程度,不仅能获得较好的模拟结果,而且可在一定程度上从微观机理解释宏观居住分布模式,是对传统统计模型的有益补充。     

基于多智能体强化学习的居住用地增长模拟——以深圳市为例

城市空间增长模拟对于土地快速城市化背景下空间规划的制定具有重要意义,在基于多智能体的城市空间增长模拟模型中,个体空间决策的建模至关重要。该文在基于效用函数的个体空间决策模型中,引入扩展的强化学习模型表征个体学习行为以对个体空间决策模型进行修正,结合政府规划因素构建耦合强化学习的居住用地增长多智能体模拟模型,模拟深圳市2005-2015年居住用地增长情况。结果表明:1)相比无学习模型,引入扩展强化学习模型的模拟结果精度提高了10.8%,更贴合居民个体在空间决策中的行为;2)引入强化学习的模拟结果中,新增居住用地在空间上呈现较强的集聚性,表明引入强化学习后模型不仅能够反映位于已有居住用地中间的填充式增长,而且能够发现期间新出现的连片居住用地;3)学习模型的参数敏感性实验结果显示,学习强度对模拟精度影响较大,表明居民个体空间决策建模中,对于他人经验的学习过程较为重要,应给予充分考虑。引入强化学习的模型有助于揭示城市增长中个体行为模式,为制定城市空间规划提供参考。     

基于多智能体建模技术的城市用地时空演化研究

为应对现代城市发展在用地资源分配和生态环境保护上遇到的巨大挑战,研究在综合复杂系统理论、城市经济学以及地理信息系统等多种学科技术的前提下,应用多智能体建模软件NetLogo构建了一个以“由下至上”运行机制为主的城市用地拓展模型,通过不同微观个体之间的互动反馈行为来实现对圣地亚哥城市宏观用地布局的模拟,并应用逐点比对和整体比对两种不同的精度检验方法来测算模型分析结果的准确性;针对当前模型研究存在的问题和未来可能的改进方向进行了归纳分析。     

基于CA模型的乌鲁木齐都市圈城市用地扩展模拟研究

借助CA模型技术,基于乌鲁木齐都市圈城市用地发展的特点,构建了模拟多个城市群体扩展的CA模型,重点对模型的转换规则进行了扩展,把用地适宜性约束、城市内部社会经济驱动、城市之间相互作用力、区域生态格局限制、城市发展规模有限这几个方面的作用机制融入到模型的规则制定中,通过宏观约束、中观调节、微观驱动3个层次逐步实现;为了使社会经济数据与CA模型中的空间数据相匹配,对社会经济数据进行了空间化表达;在历史数据的参照下,所构建的模型经过反复调试与修正,模拟结果达到了较为满意的效果;应用此模型模拟了经济优先、生态优先与规划优先3种不同情景下的城市用地扩展,对都市圈城市用地今后的发展起到重要的启示作用。     

职住空间关系研究的比较述评与展望

职住空间关系因其丰富的内涵和复杂的要素关系而成为多学科研究的热点问题。虽然跨学科交叉融合研究是职住关系研究的发展方向,但由于学科背景的差异各学科关注了职住关系问题的不同方面,其研究立场和结论也存在一定差异。通过梳理职住关系相关问题研究的主要内容、观点以及存在的争议,从学科背景这个全新的视角探寻了研究结论差异性的根源,进而展望了未来可能的研究方向和领域。分析发现：社会学侧重于社会政策和个体属性差异影响下的弱势群体职住空间错配研究,对“职住空间错配”假说的存在有较大争议;城市经济学强调市场对职住空间关系的决定性作用,认为职住分离是城市发展的必然阶段;城市规划学者试图通过测度城市通勤验证职住平衡的有效性,虽然备受争议,但其仍是城市规划的理想模型和指导思想;城市地理学则从宏观的空间测度和微观的个体行为变化洞悉城市空间重构及职住分离的影响因素和形成机制。总结认为,地理学者应在借鉴多学科的理论和方法的基础上,宏观方面要加强基于大数据和全要素信息的职住空间关系集成模型构建和精细化模拟研究,微观上,基于后现代主义和人本主义的职住关系对个体或特殊群体的影响及其响应的系统研究具有较大发展空间。     

基于细胞自动机与多主体系统理论的城市模拟原型模型

文章从城市地理学模拟模型研究发展相对滞后的现实出发,分析了传统城市模型模拟存在的问题与不足,讨论了计算机科学、复杂性研究、地理信息科学与技术、新发展的地学计算方法等作为计算城市模型发展基础的可行性,提出了一个基于细胞自动机与多主体系统理论与方法、包容了多尺度(宏观、中观、微观)层次的综合可计算城市模拟原型模型框架,对以邻里社区为基础的居住区位微观模拟模型作了概念上的讨论,分析了地学计算方法在城市模拟模型研究中的发展前沿。     

生态阻力面模型构建及其在城市扩展模拟中的应用

本文将城市扩展视为城市用地克服生态阻力向外扩散的过程,基于最小累积阻力模型(MCR)进行方法创新,在模型中引入不同等级源的相对阻力因子,并考虑生态障碍对于城市扩展的刚性约束,构建适合于城市扩展模拟的生态阻力面模型(UEER)。在此基础上,利用广州市土地利用第二次调查数据、遥感影像数据、DEM数据以及其他生态要素相关数据,通过源的确定与分级、基面阻力综合评价、UEER模型运算等步骤,构建了广州市城市扩展的生态阻力面,并用于模拟城市用地扩展至不同规模情景时的空间分布及边界。结果表明:①通过UEER模型生成的生态阻力面能够综合反映城市扩展水平过程所需克服的生态阻力,因此能够反映生态约束下城市扩展的空间运动趋势,可以用于城市扩展模拟。②与基于MCR模型的模拟结果相比,基于UEER模型的模拟结果更加符合实际并体现生态保护的要求。从城市扩展的规模与强度控制看,模拟结果更加符合实际需求,并体现政策调控方向。从城市形态以及与生态要素的关系看,一些重要的生态要素在快速城市化进程中能够得以保留,同时生态障碍作为生态隔离,能够有效地防止城市的蔓延式扩展,从而使城市扩展表现出明显的组团式特征。     

基于多智能体系统的城市增长时空动态模拟——以江苏省连云港市为例

城市化是土地利用/覆盖变化中最典型形式之一,探索城市增长的驱动机制并预测其未来变化,对于实现城市可持续发展十分重要。鉴于多智能体系统强大的模拟复杂空间系统的能力,基于联合"自上而下"和"自下而上"决策行为的视角,构建了一个城市增长时空动态模拟多智能体模型,在模型中,宏观Agent实施的"自上而下"的宏观土地利用规划行为和微观Agent自主发起的"自下而上"的微观土地利用空间诉求行为通过二维空间网格相互作用,并通过联合决策共同推动研究区域的城市化进程。以连云港市中心城区为例,考虑了基于目前趋势、经济发展优先和环境保护优先的3种目标情景,并进行了相应的城市增长情景模拟。模拟结果表明:联合"自上而下"和"自下而上"决策行为的城市增长时空动态模拟多智能体模型能够充分发挥多智能体系统的潜力来了解城市化的驱动机制,为城市管理提供基于情景分析的决策支持。     

基于遗传算法自动获取CA模型的参数

本文提出了基于遗传算法来寻找CA模型最佳参数的方法。CA被越来越多地应用于城市和土地利用等复杂系统的动态模拟。CA模型中变量的参数值对模拟结果有非常重要的影响。如何获取理想的参数值是模型的关键。传统的逻辑回归模型运算简单,常常用来获取模型的参数值,要求解释变量间线性无关,所以获取的城市CA模型参数具有一定的局限性。遗传算法在参数优化组合、快速搜索参数值方面有很大的优势。本文利用遗传算法来自动获取优化的CA模型参数值,并获得了纠正后的CA模型。将该模型应用于东莞1988~2004年的城市发展的模拟中,得到了较好的效果。研究结果表明,遗传算法可以有效地自动获取CA模型的参数,其模拟的结果要比传统的逻辑回归校正的CA模型模拟精度高。     

Forthcoming events

Abstract

This is the second of two papers which elaborates a framework for embedding urban models within GIS. In the first paper (Batty and Xie 1994), we outlined how the display functions of a proprietary GIS could be used to organize a series of external software modules which contained the central elements of the modelling process, namely dataset selection and analysis, and model specification, calibration, and prediction. In that paper, we dwelt on display and data analysis functions whereas here we outline the model-based functions of the system. We begin by reviewing residential location models based on population density theory, stating continuous and discrete model forms, and calibration methods. We then illustrate a pass through the software using data for the Buffalo urban region, showing how observed data and model estimates can be evaluated through graphic display. We present ways in which the system can be used to explore and fit a variety of models to different zoning systems and in so doing, show how subset selection and aggregation can be used to find models with good fit. Finally we draw conclusions and outline an agenda for further research.     

Discriminant analysis of the geomorphic characteristics and stability of landslide dams

Landslides can cause the formation of dams, but these dams often fail soon after lake formation. Thus, rapidly evaluating the stability of a landslide dam is crucial for effective hazard mitigation. This study utilizes discriminant analysis based on a Japanese dataset consisting of 43 well documented landslide dams to determine the significant variables, including log-transformed peak flow (or catchment area), and log-transformed dam height, width and length in hierarchical order, which affect the stability of a landslide dam. The high overall prediction power (88.4% of the 43 training cases are correctly classified) and the high cross-validation accuracy (86%) demonstrate the robustness of the proposed discriminant models PHWL (with variables including log-transformed peak flow, and log-transformed dam height, width and length) and AHWL (with variables including log-transformed catchment area, and log-transformed dam height, width and length). Compared to a previously proposed “DBI” index-based graphic approach, the discriminant model AHV – which uses the log-transformed catchment area, dam height, and dam volume as relevant variables – shows better ability to evaluate the stability of landslide dams. Although these discriminant models are established using a Japanese dataset only, the present multivariate statistical approach can be applied for an expanded dataset without any difficulty when more completely documented worldwide landslide-dam data are available.     

基于互联网数据的城市社区租金评估及 空间格局制图——以深圳市为例

准确刻画精细化尺度下的城市房租空间格局,对于研究城市居住行为、提高城市规划合理性十分重要。文章提出了一种基于互联网房租数据作为可靠数据源的城市房租空间格局制图方法。以深圳市作为研究区,通过广泛采集开放平台中用户发布的租房信息,绘制社区行政区尺度的房租空间分布图。房租空间制图涉及到对于没有样本数据区域平均房租的评估,因此,选取一系列与房租相关的房产属性、房屋区位及配套设施的评价指标,采用前馈神经网络技术构建评估模型。以2015年深圳市的住宅租赁市场作为研究对象,通过对结果的分析,以及与权威部门发布的统计数据进行比较,表明此方法能够有效地绘制社区尺度下城市房租的空间分布,模型预测结果的误差（%RMSE）为13.87%。所使用的互联网房租数据、POIs数据及前馈神经网络的建模工具均是开源的,而且所提出的方法论具有普适性,能够应用于其他研究区的房租空间格局制图,具有实践意义。     

Rapid three-dimensional urban model production using bilayered displacement mapping

The depiction and navigation of large-scale urban landscapes are limited by the great cost of traditional computer-aided design (CAD) models for large urban environment in terms of both the labor of data entry and the runtime computational expense. This article presents a hybrid modeling approach that enables rapid urban model production from legacy spatial data. Our scheme fills the gap between the low geometry models, such as photo-textured digital terrains, and high geometry models, such as true three-dimensional CAD models. To achieve optimal performance in modeling and rendering, we employ bilayered displacement mapping consisting of global displacement mapping (GDM) for terrains and local displacement mapping (LDM) for buildings. The LDM is performed only within image processing so that the complexity of the models depends only on the area of an urban model. We present a use case of rapid urban model production to compare our approach with the traditional polygonal urban models of a widely used geo-browser.     

Not developed yet? Alternative ways to include locations without changes in land use change models

Biogeographers have developed a new generation of statistical models called presence-only models, which require no data concerning the absence of a species and do not assume that the absence of a species indicates habitat unsuitability. Both characteristics are especially useful when modeling a species that is actively spreading across a landscape. Although urban expansion is sometimes equated to an invading species, the applicability of presence-only models has not yet been explored when modeling urban growth. This article compares predictions of urban growth using a presence-only model (ecological niche factor analysis) and a more traditional presence–absence model (logistic regression). An additional model used pseudo-absence sites, from the presence-only model output, as input into the presence–absence model. The models were applied to New Jersey's Barnegat Bay Watershed. Overall, the traditional presence–absence model performed the best, although the presence-only model was sufficiently similar to warrant further exploration of presence-only models when no reliable absence data (i.e., locations where no conversion occurred) exists. However, due to data-formatting requirements of the presence-only model, it is difficult to accommodate data pertaining to administrative boundaries, which are inherently Boolean. Finally, the output based on the pseudo-absence approach overpredicted urban conversion when compared to the other approaches.     

Measuring short-term population mobility among indigenous Australians: options and implications

"Despite the fact that indigenous Australians are known to be frequently mobile over the short term, statistical information regarding this population movement is grossly deficient.... This paper examines various means by which short-term population movement can be quantified to yield aggregate indicators of demographic impact. First, census data are used to establish the rate and pattern of inter-regional, short-term displacement. This reveals regional urban centres as net recipients of temporary residents while most rural areas experience temporary absenteeism. Secondly, results from household surveys are reported stressing the importance of including visitors to households in the estimation of service populations. Thirdly, administrative data on occupancy in urban hostels are used to derive indicators of the duration of movement."     

Multi-level temporal autoregressive modelling of daily activity satisfaction using GPS-integrated activity diary data

In this research, we match web-based activity diary data with daily mobility information recorded by GPS trackers for a sample of 709 residents in a 7-day survey in Beijing in 2012 to investigate activity satisfaction. Given the complications arising from the irregular time intervals of GPS-integrated diary data and the associated complex dependency structure, a direct application of standard (spatial) panel data econometric approaches is inappropriate. This study develops a multi-level temporal autoregressive modelling approach to analyse such data, which conceptualises time as continuous and examines sequential correlations via a time or space-time weights matrix. Moreover, we manage to simultaneously model individual heterogeneity through the inclusion of individual random effects, which can be treated flexibly either as independent or dependent. Bayesian Markov chain Monte Carlo (MCMC) algorithms are developed for model implementation. Positive sequential correlations and individual heterogeneity effects are both found to be statistically significant. Geographical contextual characteristics of sites where activities take place are significantly associated with daily activity satisfaction, controlling for a range of situational characteristics and individual socio-demographic attributes. Apart from the conceivable urban planning and development implications of our study, we demonstrate a novel statistical methodology for analysing semantic GPS trajectory data in general.     

Calibrating a Land Parcel Cellular Automaton (LP-CA) for urban growth simulation based on ensemble learning

The reliability of raster cellular automaton (CA) models for fine-scale land change simulations has been increasingly questioned, because regular pixels/grids cannot precisely represent irregular geographical entities and their interactions. Vector CA models can address these deficiencies due to the ability of the vector data structure to represent realistic urban entities. This study presents a new land parcel cellular automaton (LP-CA) model for simulating urban land changes. The innovation of this model is the use of ensemble learning method for automatic calibration. The proposed model is applied in Shenzhen, China. The experimental results indicate that bagging-Naïve Bayes yields the highest calibration accuracy among a set of selected classifiers. The assessment of neighborhood sensitivity suggests that the LP-CA model achieves the highest simulation accuracy with neighbor radius r = 2. The calibrated LP-CA is used to project future urban land use changes in Shenzhen, and the results are found to be consistent with those specified in the official city plan.