Algorithmic landscapes : computational methods for the mediation of form, information, and performance in landscape architecture

Abstract

Recent design discourse in landscape architecture has shown a prejudice against rigid formal systems and has advocated systems that promote self-organization, emergence, and indeterminacy, as well as non-linear systems and thinking. A growing number of landscape architectural practices have seized upon the thinking and tools associated with dynamic systems in order not only to create new types of formal expression, but in the hope that more performative landscapes can be conceived and implemented. The role of the designer in such projects shifts from that of the artist who gives concrete form and definition to a project, to the engineer of a system, who sets in motion a process or series of processes that will ultimately lead to a range of possible futures and associated forms. The development of algorithms, in the form of scripts and codes, presents a way for the designer to more rigorously study the behavior of complex systems and the formal, spatial, and evolutionary implications of dynamic processes. Already, disciplines at the edges of landscape architecture, from the natural science such as geomorphology and botany, to the applied sciences such as environmental engineering and planning, have expanded their thinking through algorithmic simulations and models, in order to answer fundamental questions concerning the behavior of dynamic systems and the emergence of form in nature. Computer scientists have been particularly interested in an algorithmic description of nature for ends ranging from improving graphical representation, to generating believable artificial worlds, to improving artificial intelligence. For landscape architects and designers, applying such algorithmic processes and thinking allow the designer to use the computer not only as a tool of production and representation, but as a potential tool for analysis and design. While the holistic approach of Landscape architects should not require the same training and approach as that of the engineers and coders of complex systems, in order to fulfill the high ambitions of recent design discourse in a meaningful way, landscape architects should reassess their relationship with computational processes and tools and engage them in a new way. As such, this thesis engages the role of algorithms in landscape design in terms of three fundamental categories. First, the role of algorithms in generating form in the landscape. Second, how information can shape and enrich algorithms, and how algorithms can mediate between complex data sets. Third, the role of performance in the contemporary landscape and how algorithms can be used to make landscapes that respond better to change through time and which better reflect underlying processes. The categories of algorithmic form, information, and performance are explored in this thesis through three lenses and in three parts. Part I introduces algorithms and algorithmic design in the context of landscape architecture, and then provides a historical underpinning describing how formal systems used in design contexts emerged and later developed through various periods of human history, ultimately being automated through computation. Formal systems did not emerge in a vacuum and they are useful only insofar as they mediate between layers of natural and cultural information and describe the structures and patterns observed in nature and in culture. These observed structures and patterns are explored in more detail through specific algorithmic patterns and paradigms in Part II of the thesis. In Part III, the patterns are combined into specific design case-studies of landscape architectural projects developed by the author. The historical background in Part I, the computational experiments of Part II, and the projects presented in Part III contribute to an overall picture showing the potentials and difficulties with adopting algorithmic methods to simulate, model, and design landscapes in complex contemporary contexts.