From Zurich to New York: Exploring Programmatic Settlement Development Across Continents
FEATURE
By Yves Räth
Yves Räth is a PhD student at the Chair of Planning of Landscape and Urban Systems at the Swiss Federal Institute of Technology (ETH Zurich). He joined the Urban Systems Lab at The New School as a Visiting Scholar in winter and spring 2024. He holds a BS in Geography and a MS in Spatial Development and Infrastructure Systems. Yves' research focuses on historical regional settlement development, with an emphasis on emergent patterns of urban form. At the Urban Systems Lab, he identified settlement development patterns that can be used to help urban planners make decisions through a deeper understanding of predicted long-term development dynamics.
KEY TAKEAWAYS
My research on Swiss settlement patterns using machine learning faced challenges when applied to New York City, revealing the critical importance of building height in defining urban neighborhoods in different contexts.
Contrasting Swiss and American urban landscapes highlighted the need for adaptable methodologies in my work, as models effective in one cultural setting may not directly translate to another.
My experience at the Urban Systems Lab and the IALE conference broadened my research perspective, encouraging me to consider social justice aspects alongside my quantitative analysis of urban transformation.
Observing diverse cities like New York and Oklahoma City provided valuable insights into varied urban development patterns, reinforcing the importance of context-specific approaches in my analysis of urban form and evolution.
In my research, I investigate archetypal developments of settlements on the Swiss Plateau (Figure 1). My arrival to New York City prompted me to think about the potential applicability of my research in the American context.
Figure 1: The Swiss Plateau.Source: Yves Räth, 2024, Basemap: swisstopo (2023)
From the Swiss Plateau to Manhattan: Contrasting Urban Landscapes
Over the past 120 years, settlements in Switzerland have undergone significant development, particularly on the Swiss Plateau. This region has become one of Europe's most densely populated areas, featuring a polycentric network of hamlets, villages, and major cities like Zurich, Bern, and Lausanne. Despite considerable transformation and expansion, Most settlements have preserved their historic cores, which reflect medieval urban planning designed for pedestrian transportation. From a programmatic perspective, these cores typically remain high-density, mixed-use neighborhoods, blending residential and commercial activities. This finding informed my interest in the relationship between urban form and its program as well as its evolution over time.
As settlements grew, they diversified into distinct programmatic zones. Low-density residential areas emerged alongside separate commercial and industrial districts. Industrial zones, due to their high emissions (noise, pollution, etc.), evolved into larger, more isolated neighborhoods (Figure 2). This transformation raises a key question: How have the sizes of different neighborhood types changed over time in various locations? The evolution of a programmatic neighborhood’s share of the settlement’s total area can inform a development pattern type. As they consider the potential future growth of settlement, development patterns could provide valuable insights for contemporary urban planners. My research aims to explore these changes and uncover any common trajectories in settlement development. I seek to understand these trends to anticipate future settlement evolution and inform urban planning strategies.
Figure 2: Example of a settlement with different types of urban form, correlated with residential (Res.), commercial and industrial (Com. & Ind.), and mixed-use (Mix.) uses, as well as low (L), medium (M), and high (H) densities. Source: Yves Räth, 2024
Urban form serves as my proxy for land use in defining neighborhood types. I analyze buildings’ orientation, size, distribution, and shape and measure their similarity to neighboring buildings. These metrics, analyzed over time using historic maps, enable neighborhood classification using machine learning algorithms.
My historical algorithm is trained using maps dating to the 1870s, produced every 10-30 years for military purposes across Switzerland (Figure 3). Classification is informed by the individual building footprints provided by these maps. But height information is lacking altogether. In the Swiss context, this omission is negligible; neighborhood types are identifiable primarily through building footprint area, orientation variance, and inter-building distance. However, this methodology's applicability to New York City, particularly Manhattan—where I lived and worked for over six months—intrigued me. The big apple’s iconic third dimension (e.g. skyscrapers) and distinctive urban form characterized by its rigorous street grid, present a compelling evolution leaving me curious about its impact on programmatic neighborhood types.
Figure 3: Historical Map of Altstetten in the year 1870. Individual buildings can be distinguished as well as other features. © swisstopo
Methodology in Motion: Adapting Research Techniques Across Continents
When examining building footprints through maps, the area of Manhattan south of Wall Street appears to have retained its historic urban fabric, in parts resembling a European street layout. However, a consideration of building heights reveals a starkly different reality: most structures have been progressively replaced by ever-taller high-rises and skyscrapers. Concurrently, these neighborhoods have undergone a programmatic shift from mixed-use to almost exclusively commercial. This transformation is partially visible on two-dimensional maps, evidenced by the increasing merger of building footprints—from smaller walk-ups to the larger footprints of high-rises. Nevertheless, building height emerges as a critical feature of urban form, necessary for defining neighborhood types in New York City. This observation underscores the importance of context: a model of analysis that performs well in one cultural setting may not be equally effective in another. This realization motivates the improvement of the current framework by incorporating building height into the machine learning algorithm, aiming to better predict and characterize neighborhood types.
Throughout my time in New York City, I continued to study settlement development patterns in the Swiss Plateau. I labored over training the machine learning algorithm and could identify specific development pattern types. Over time, some settlements primarily transform their program in low-density neighborhoods, while other settlements show an exponential programmatic densification. Another typical development pattern is the "high-density developer" where a settlement’s mixture of neighborhoods significantly increase their share of high-density mixed-use, commercial, and industrial areas. New York is a prime example of this development pattern type.
My research method, originally designed for the Swiss Plateau, proved challenging to apply in New York City. However, my time at the Urban Systems Lab yielded valuable methodological insights and fostered a productive, inspiring period in my research. This experience allowed me to explore new dimensions of urban transformation and critically examine the concepts of stability and change. The connections I made in the lab were equally valuable. Living and working in such a dense, vibrant city exposed me to diverse perspectives. Regular lab meetings provided intellectual stimulation, featuring contributions from researchers and guests across various fields. The diverse cultural and academic backgrounds of my collaborators served as a constant source of inspiration. While my work primarily focuses on the quantitative assessment of change, the lab's emphasis on social justice in the context of urban transformation highlighted potential new avenues for future research. This interdisciplinary exposure broadened my perspective, encouraging me to consider the human dimensions of urban change alongside quantitative measures.
Source: Yves Räth, 2024
New Perspectives: Insights from the Urban Systems Lab and Beyond
A highlight of my North American visit was presenting at the International Association for Landscape Ecology (IALE) North America Annual Meeting in Oklahoma City. My paper, "Tracing urban transformations from historical maps," allowed me to refine my interim findings and reassess my research direction in collaboration with the Urban Systems Lab. The conference fostered valuable discussions, new friendships, and fresh inspiration for my work.
Oklahoma City offered a stark contrast to both New York and Swiss settlements, revealing a different facet of American urbanism. Its dense yet nighttime-deserted commercial core (Figure 4) juxtaposed against sprawling suburbs epitomizes car-centric urban design. Recently revitalized areas like Bricktown (Figure 5) seek to break this paradigm with a pedestrian mall buzzing with restaurants and bars. The city's distinct urban patterns make it an ideal case study for neighborhood classification using two-dimensional map features. Suburban areas, characterized by winding roads and widely-spaced smaller buildings, differ markedly from the commercial center's large footprints housing skyscrapers. This clear differentiation suggests that a map-based classification approach would be particularly effective here.
Figure 4: Oklahoma City, OK. Source: Yves Räth, 2024
Figure 5: Bricktown, Oklahoma City. Source: Yves Räth, 2024
As I reflect on my journey from the Swiss Plateau to New York City and Oklahoma City, I am struck by the diverse urban landscapes I've encountered and the valuable insights gained. This experience has reinforced the importance of context in urban studies and the need for adaptable methodologies. While my research began with a focus on Swiss settlement patterns, the contrasts observed in American cities have broadened my perspective on urban form and development. The challenges faced in applying Swiss-based models to New York's unique vertical landscape highlight the complexity of urban systems and the need for nuanced, context-specific approaches. Moving forward, I am eager to incorporate these new insights into my research, particularly in considering building height and cultural factors in urban pattern analysis. This cross-continental exploration has not only enriched my academic work but also deepened my appreciation for the dynamic nature of urban development. As cities continue to evolve, our understanding of their patterns and processes must evolve as well, adapting to new contexts and challenges in the ever-changing urban landscape.
