Research PaperSpatial statistical analysis and simulation of the urban heat island in high-density central cities
Graphical abstract
Introduction
Projections by the United Nations suggest that 60% of the world's population will reside in urban regions by 2030. The resulting expansion of impervious surfaces is likely to intensify the urban heat island (UHI), whereby temperatures in urban core areas are higher than in surrounding suburban and rural areas. The UHI induces heat stress, tropospheric ozone formation, and resulting health problems. Higher temperatures lead to increased electricity demand for air conditioning, which, in turn, raises power plant pollution and greenhouse gas emissions. In addition, the UHI may increase water temperatures, resulting in water ecosystems impairment. It is clear that mitigating UHI impacts requires comprehensive planning strategies accounting for the effects of urban morphology, infrastructure, and greening, on the UHI. However, a lack of understanding of these effects has been a primary obstacle to implementing such strategies.
The objectives of this paper are to: (1) specify and estimate novel statistical regression models that account for both spatial neighborhood effects and the simultaneous effects of several urban characteristics on land surface temperatures derived from satellite imagery; (2) assess which grid scale used to capture the data yields the statistical model with the best explanatory and predictive power; and (3) use the best regression model to numerically simulate UHI-mitigating strategies in an urban design and planning context. To model the statistical relationship, both two-dimensional (2-D) and three-dimensional (3-D) information is used to represent the complex geometric structure of urban centers, with an application to the urban core of Columbus, OH. Earlier UHI research has primarily used 2-D information, such as land uses delineated with satellite imagery and building ground floor boundaries produced by geographic information systems (GIS). In the case of homogeneous land uses, this data may be sufficient to predict surface temperatures (Carlson & Arthur, 2000). However, 3-D information is necessary to analyze more complex sites, including dense building clusters (Unger, 2006, Unger et al., 2004). LiDAR data are used to generate 3-D urban geometry characteristics. A hierarchy of grids, with cell sizes of 480 m, 240 m, and 120 m, is used to integrate all the data. A spatial regression model capturing neighborhood effects and the relationship between surface temperatures and the geometric characteristics and other factors of the built environment is formulated and estimated for these different grids. The best model is then used to simulate different greening scenarios for the center of Columbus, illustrating its potential for mitigating the UHI.
The remainder of the paper is organized as follows. Section 2 consists in a review of the relevant literature. Section 3 describes the study area, data sources, and data processing. Regression models are formulated and estimated in Section 4. The simulations of greening scenarios are presented in Section 5. Section 6 discusses remaining issues and areas for further research. Section 7 concludes the paper.
Section snippets
Literature review
There has been an explosive growth in the research literature on the UHI in recent years. Using the search engine Google Scholar and typing in “urban heat island” returns 25,300 distinct journal articles, book chapters, books, working papers, and professional reports, as of October 9, 2013, with 86.5% of these works produced since 2001. Basic reviews of research on the energetic basis of the UHI and on satellite-derived thermal remote sensing of urban areas can be found in Oke (1982), Roth,
Study area, data sources, and data processing
This research focuses on a densely built section of the City of Columbus, OH, an area of 46.5 km2 including the central business district (CBD). The Scioto and Olentangy rivers flow from north to south, and merge into one channel near the CBD. There are parks and outdoor recreation sites in the southern part of the study area, which also includes residential areas. A continuous population growth since 1950 has led to much building construction, resulting in increased surface temperatures by at
Model estimation
Regression analysis is used to explore the relationship between LST and the urban characteristics discussed in Section 3 over each grid structure, and to determine the estimation method that best represents the UHI, including the ordinary least squares (OLS) model, the spatial lag (autoregressive) model (SAR), and the general spatial model (GSM), a combination of the SAR and SEM (spatial error model) models. The temperature observed at a given location (cell) is likely to be spatially
Simulation of greening scenarios
What are the impacts of alternative greening actions on surface temperatures? To answer this question, the 120 m GSM model presented in Table 3 is used in a simulation mode, computing the temperature of each cell under a specific greening scenario. Before describing the proposed scenarios, the basic computational steps of the simulation are first presented. Assuming that the error terms u and ɛ are equal to 0, Eq. (13) can be rewritten in matrix form as:
Discussion
Several issues related to the proposed methodology and its simulation implementation are first discussed, including the relationship between surface and air temperatures, the role of water, the measurement of vegetation, and the assessment of UHI-mitigating strategies. Extensions of the research approach are next outlined.
What is the relationship between satellite-derived surface temperature and air temperature, which most directly affect people? Based on their review of the available
Conclusions
The objectives of this research were to specify and estimate improved statistical regression models of the UHI, determine the best scale of capture of the necessary data, and illustrate the use of the model(s) for designing UHI-mitigating strategies. These objectives have been, by and large, achieved.
Satellite data have been used to estimate surface temperatures and vegetation coverage, GIS data to delineate building boundaries and land uses, and LiDAR data to closely approximating the real
Acknowledgments
The comments and suggestions of the Associate Editor and three referees on an earlier draft are very much appreciated.
B. Chun is a Post-Doctoral Researcher at the Center for Geographic Information Systems, and Lecturer in City and Regional Planning, Georgia Institute of Technology. He received his PhD in City and Regional Planning from The Ohio State University in 2012. His research interests focuses on three-dimensional urban modeling, GIS, energy and climate change. He has published several articles on these issues.
References (44)
- et al.
Spatial–temporal dynamics of land surface temperature in relation to fractional vegetation cover and land use/cover in the Tabriz urban area, Iran
Remote Sensing of Environment
(2009) - et al.
Reductions in air conditioning energy caused by a nearby park
Energy and Buildings
(1998) - et al.
The impact of land use-land cover changes due to urbanization on surface microclimate and hydrology: A satellite perspective
Global and Planetary Change
(2000) - et al.
On the relation between NDVI, fractional vegetation cover, and leaf area index
Remote Sensing of Environment
(1997) - et al.
Remote sensing image-based analysis of the relationship between urban heat island and land use/cover changes
Remote Sensing of Environment
(2006) Urban nocturnal temperatures, street geometry and land use
Atmospheric Environment
(1996)- et al.
Daytime urban heat island effect in high-rise and high-density residential developments in Hong Kong
Energy and Buildings
(2004) Influence of a large urban park on temperature and convective precipitation in a tropical city
Energy and Buildings
(1990)- et al.
Estimation of storage heat flux in an urban area using ASTER data
Remote Sensing of Environment
(2007) - et al.
Urban tree transpiration over turf and asphalt surfaces
Atmospheric Environment
(1998)
Temporal and spatial assessments of minimum air temperature using satellite surface temperature measurements in Massachusetts, USA
Science of the Total Environment
Investigating spatial non-stationary and scale-dependent relationships between urban surface temperature and environmental factors using geographically weighted regression
Environmental Modeling & Software
Street design and urban canopy layer climate
Energy and Building
Vegetation as a climatic component in the design of, an urban street: An empirical model for predicting the cooling effect of urban areas, with trees
Energy and Buildings
Downscaling AVHRR land surface temperatures for, improved surface urban heat island intensity estimation
Remote Sensing of Environment
Thermal remote sensing of urban climates
Remote Sensing of Environment
Comparison of impervious surface area and normalized difference vegetation index as indicators of surface urban heat island effects in Landsat imagery
Remote Sensing of Environment
Spatial pattern of impervious surfaces and their impacts on land surface temperature in Beijing, China
Journal of Environmental Sciences
Spatial econometrics: Methods and models
Two decades of urban climate research: A review of turbulence, exchanges of energy and water, and the urban heat island
International Journal of Climatology
Canyon geometry, street temperatures and urban heat island in Malmö, Sweden
Journal of Climatology
A multiple linear statistical model for estimating the mean maximum urban heat island
Theoretical and Applied Climatology
Cited by (265)
A new framework quantifying the effect of morphological features on urban temperatures
2023, Sustainable Cities and SocietyThermal and humidification effects of a swimming pool in Hong Kong's humid-subtropical summer
2023, Sustainable Cities and Society
B. Chun is a Post-Doctoral Researcher at the Center for Geographic Information Systems, and Lecturer in City and Regional Planning, Georgia Institute of Technology. He received his PhD in City and Regional Planning from The Ohio State University in 2012. His research interests focuses on three-dimensional urban modeling, GIS, energy and climate change. He has published several articles on these issues.
J.-M. Guldmann is Professor Emeritus of City and Regional Planning at The Ohio State University. He holds a PhD in Urban and Regional Planning from the Israel Institute of Technology, Haifa, Israel, and a Masters in Industrial Engineering from the Ecole des Mines, Nancy, France. His research focuses on the quantitative modeling of urban structure, energy and telecommunications infrastructures, and environmental planning, particularly air quality and water issues. He has published a book and over 70 articles and book chapters on these issues.