A Growing Urban Problem Frank Schweitzer Urban growth is one of the biggest challenges for humankind in the twenty-first century. The world's urban population is at present estimated to be growing by about 50 million people per year. This growth is still almost exponential, and saturation may be reached only by the middle of the century. So there is much discussion about estimates of land consumption, the manageable size of megacities, and simple ways to structure, if not to control, this vast spread of urbanization. For this, a better understanding of the spatial dynamics of urban growth should be a precondition. In fact, urban planning, urban geography, urban economics and other related disciplines have put quite some effort into this problem. Although they provide insight into some details, they do not present the big picture of generic mechanisms underlying these complex dynamics. So it is reasonable to consider whether the theory of complex systems developed in different scientific areas over the past 30 years may provide a suitable toolbox for gaining such an insight. In Cities and Complexity , Mike Batty, an expert in the area of urban modelling, unfolds the different aspects of urban change in relation to complex systems theory. The focus is on models and computer simulations: there is no introduction explaining what we already know about cities and growth, or any attempt at orientation on the phenomenological side. This is good and bad at the same time. Some readers will certainly miss a clear relation between cities as we know them and the rather abstract models and results for specific dynamic features, such as segregation or polynucleated growth. Others, however, will appreciate the insights into generic features of urban dynamics that can only be highlighted at the abstract level used. As the author puts it, the focus is “largely on experiments with models that provide us with analogies as to how cities develop and evolve”. These analogies are investigated on very different modelling levels relating to the different scales of urban dynamics. Models of fractal growth show analogies to the morphological structure of urban aggregates, and models of moving agents resemble the motion of pedestrians on streets and inside buildings. Many of these models use cellular automata, which allow a broad range of topics such as land-use patterns or regeneration of urban areas to be addressed. A ‘desktop simulator' is described in more detail in one of the chapters, tempting the reader to play with the ideas discussed in the book. Unfortunately, the plan to put examples from the book together with related material on a website (www.complexcity.info) is yet to be realized. The book certainly succeeds in making the topic of urban growth accessible and interesting to those already familiar with formal modelling of complex systems. However, practitioners in urban planning may find some of the concepts and models too abstract and ‘ academic' to give solutions. But at least they should agree that this complex-systems perspective sheds new light on to the dynamics of urban evolution by highlighting relations to seemingly distant phenomena, such as swarming or epidemics. Apart from the necessary technical details of the models, the author gives a number of illustrative examples and facts to elucidate his viewpoint, widening the appeal of the book to a broader audience. Frank Schweitzer is chair of systems design at ETH Zürich, Switzerland .

日益增长的城市问题 弗兰克·施韦策 城市增长是二十一世纪人类最大的挑战之一。根据目前的预测，全世界的城市人口每年增长大约五千万。这种增长几乎是指数式的，在本世纪中期就可能达到饱和。因此就有了很多讨论和评估，关于土地消耗、特大城市可管理的面积以及组织而非控制这种快速传播的城市化的简单方法。 为此，更好地理解城市增长的空间动力学应该是一个先决条件。实际上，城市规划学、城市地理学、城市经济学以及其它的相关学科已经在这个问题上作了很大的努力。虽然它们对一些细节有了深入研究，但是没有提出蕴含在这些复杂动力学下面的总的机制的大框图。因此有理由考虑是否过去三十年来在不同科学领域所发展的复杂系统理论能够为获得这种深入见解提供一个合适的工具箱。 在《城市和复杂性》中，城市建模领域的专家迈克·贝蒂揭示了与复杂系统理论相关的城市变化的不同方面。焦点集中于模型和计算机模拟：没有对于我们已知的城市和增长或者现象学方向的任何尝试的解释性介绍，这喜忧参半。有些读者肯定会无法搞清正如我们已知的城市和对于特定动态特征（比如种族隔离和多核增长）的相当抽象的模型和结果之间的明确关系。而其他读者会欣赏对于已使用的只在抽象层面强调的城市动态学的总特征的深入见解。正如作者所说的，焦点主要在于“能提供给我们关于城市如何发展和进化的类比模型的实验”。 这些类比研究是在有关城市动力学的不同尺度的非常不同的建模层面上进行的。分型增长模型显示了城市聚集与形态学结构的相似性，而运动智能体的模型类似于街道和楼房内行人的运动。其中的许多模型使用了元胞自动机，它需要研究诸如土地使用模式和城市区域再生等广泛的课题。其中的一章很详细地介绍了一个“桌面模拟器”，引诱读者与书中讨论的观点进行互动。不幸的是，把书中的例子和相关材料一起放到网上 (www.complexcity.info) 的计划还没有实施。 这本书让那些已经熟悉以前的复杂系统建模方法的读者理解城市增长并对此话题产生兴趣无疑是成功的。但是，城市规划的实践者会发现有些概念和模型过于抽象化和“学术化”，不能解决问题。但是至少他们应该同意，通过强调与群集效应和流行病学这些表面上不相关的现象之间的关系，这种复杂系统的观点在城市进化动力学上给出了新的见解。除了一些必需的模型的技术细节，作者还给出了许多说明性的例子和事实来阐述他的观点，使得此书能够吸引更广泛的读者。 弗兰克·施韦策是瑞士苏黎世理工学院系统设计专业主任。