Classification of the heterogeneous structure of urban landscapes (STURLA) as an indicator of landscape function applied to surface temperature in New York City

Defining landscape structure and key relationships between landscape structure and function is challenging in urban areas characterized by density and patchy spatial patterns. In order to trace the spatial and temporal patterns of urban landscape structures, compare patterns across cities, or inform urban design principles, we need to classify the landscape in a way that captures context and landscape heterogeneity, but can be broadly applied across different cities or landscape variations within a city. 

This classification approach is unique in that it develops composite (as opposed to homogeneous) classes, which are defined a posteriori, based on compositions of adjacent structural elements that emerge in the urban landscape, using a cellular grid to define units of analysis. We test the separability of classes that emerge from this approach, and find that it is possible to discern classes – comprised of the mix of land and building covers common in urban areas – which have meaningfully distinct temperature signatures. This classification approach may be extended to multiple cities and ecological indicators in order to offer insight into the relationship between urban landscape structure and ecosystem response, in a way that accounts for interactions among different types of urban landscape surfaces. We suggest that this approach can support spatial prioritization of landscape function needs in urban development and design approaches for improving particular types of functioning, such as reductions in urban heat.

Zoé Hamstead, Assistant Professor
Department of Urban and Regional Planning, UB

Peleg Kremer
Timon McPhearson
Tishman Environment and Design Center, The New School

Neele Larondelle
Dagmar Haase
Humboldt University of Berlin, Institute of Geography

Ecological Indicators

2016