• Akbari, H., , L. Rose, , and H. Taha. 1999. Characterizing the fabric of the urban environment: A case study of Sacramento, California. Lawrence Berkeley National Laboratory, U.S. Department of Energy, 48 pp.

    • Search Google Scholar
    • Export Citation
  • Akbari, H., , M. Pomerantz, , and H. Taha. 2001. Cool surfaces and shade trees to reduce energy use and improve air quality in urban areas. Sol. Energy 70:295310.

    • Search Google Scholar
    • Export Citation
  • Anderson, J., , E. Hardy, , J. Roach, , and R. Witmer. 1976. A land use and land cover classification system for use with remote sensor data. Geological Survey Professional Paper 964, United States Department of the Interior–Geological Survey, United States Government Printing Office, 24 pp.

    • Search Google Scholar
    • Export Citation
  • Baatz, M. Coauthors 2004. eCognition User Guide 4. Definiens Imaging. Munchen, Germany. [Available online at http://www.gis.unbc.ca/help/software/ecognition4/ELuserguide.pdf.].

    • Search Google Scholar
    • Export Citation
  • Barnsley, M. J., , A. M. Steel, , and S. L. Barr. 2003. Determining urban land use through an analysis of the spatial composition of buildings identified in LIDAR and multispectral image data. Remotely Sensed Cities, V. Mesev, Ed., Taylor and Francis, 83–108.

    • Search Google Scholar
    • Export Citation
  • Benz, U. C., , P. Hofmann, , G. Willhauck, , I. Lingenfelder, , and M. Heynen. 2004. Multiresolution, object-oriented fuzzy analysis of remote sensing data for GIS-ready information. ISPRS J. Photogramm. Remote Sens. 58:239258.

    • Search Google Scholar
    • Export Citation
  • Blaschke, T., , C. Burnett, , and A. Pekkarinen. 2004. Image segmentation methods for object-based analysis and classification. Remote Sensing Image Analysis: Including the Spatial Domain, F. De Meer and S. de Jong, Eds., Kluver, 211–236.

    • Search Google Scholar
    • Export Citation
  • Burley, T. M. 1961. Land use or land utilization? Prof. Geographer 13:1820.

  • Dell’Acqua, F., , P. Gamba, , and G. Lisini. 2003. Improvements to urban area characterization using multitemporal and multiangle SAR images. IEEE Trans. Geosci. Remote Sens. 41:19962004.

    • Search Google Scholar
    • Export Citation
  • Donnay, J-P., , M. J. Barnsley, , and P. A. Longley. 2001. Remote sensing and urban analysis. Remote Sensing and Urban Analysis, J.-P. Donnay, M. J. Barnsley, and P. A. Longley, Eds., Taylor and Francis, 3–18.

    • Search Google Scholar
    • Export Citation
  • Federal Highway Administration cited. 2007. FHWA functional classification guidelines. (Created 1989.) [Available online at http://tpd.azdot.gov/gis/fclass/fc_fhwa_sect_2_1.php.].

    • Search Google Scholar
    • Export Citation
  • Foody, G. M. 2000. Estimation of sub-pixel land cover composition in the presence of untrained classes. Comput. Geosci. 26:469478.

  • Golden, J. S. 2004. The built environment induced urban heat island effect in rapidly urbanizing arid regions—A sustainable urban engineering complexity. Environ. Sci.: J. Integr. Environ. Res. 1:321349.

    • Search Google Scholar
    • Export Citation
  • Golden, J. S., and K. Kaloush. 2006. Mesoscale and microscale evaluations of surface pavement impacts to the urban heat island effects. Int. J. Pavement Eng. 7:3752.

    • Search Google Scholar
    • Export Citation
  • Golden, J. S., , P. Guthrie, , K. Kaloush, , and R. Britter. 2005. The summertime urban heat island hysteresis lag complexity: Applying thermodynamics, urban engineering and sustainability research. Sustainable Eng.—J. Roy. Inst. Civ. Eng. 158:197210.

    • Search Google Scholar
    • Export Citation
  • Gong, P., and P. J. Howarth. 1990. The use of structural information for improving land-cover classification accuracies at the rural-urban fringe. Photogramm. Eng. Remote Sens. 56:6773.

    • Search Google Scholar
    • Export Citation
  • Grimmond, C. S. B., , H. A. Cleugh, , and T. R. Oke. 1991. An objective urban heat storage model and its comparison with other schemes. Atmos. Environ. 25B:311326.

    • Search Google Scholar
    • Export Citation
  • Grossman-Clarke, S., , J. A. Zehnder, , W. L. Stefanov, , Y. Liu, , and M. A. Zoldak. 2005. Urban modifications in a mesoscale meteorological model and the effects on near-surface variables in an arid metropolitan region. J. Appl. Meteor. 44:12811297.

    • Search Google Scholar
    • Export Citation
  • Gui, J., , J. Carlson, , P. Phelan, , K. Kaloush, , and J. S. Golden. 2007a. Impact of pavement thickness on surface diurnal temperatures. J. Green Build. 2:121130.

    • Search Google Scholar
    • Export Citation
  • Gui, J., , P. Phelan, , K. Kaloush, , and J. Golden. 2007b. Impact of pavement thermo-physical properties on surface temperatures. J. Mater. Civ. Eng. 19:683690.

    • Search Google Scholar
    • Export Citation
  • Haack, B., , N. Bryant, , and S. Adams. 1987. An assessment of Landsat MSS and TM data for urban and near-urban land-cover digital classification. Remote Sens. Environ. 21:201213.

    • Search Google Scholar
    • Export Citation
  • Herold, M., , D. A. Roberts, , M. E. Gardner, , and P. E. Dennison. 2004. Spectrometry for urban remote sensing—Development and analysis of a spectral library from 350 to 2400 nm. Remote Sens. Environ. 91:304319.

    • Search Google Scholar
    • Export Citation
  • Jensen, J. R. 2000. Remote Sensing of the Environment: An Earth Resource Perspective. Prentice-Hall, 544 pp.

  • Jo, J., , J. S. Golden, , and S. W. Shin. 2008. Incorporating built environment factors into climate change mitigation strategies for Seoul, South Korea: A sustainable urban systems framework. Habitat Int. 33:267275.

    • Search Google Scholar
    • Export Citation
  • Kerschgens, M. J., and H. Kraus. 1990. On the energetics of the urban canopy layer. Atmos. Environ. 24B:321328.

  • Lewinski, S., and K. Zaremski. 2004. Examples of object-oriented classification performed on high-resolution satellite images. Misc. Geogr. 11:349358.

    • Search Google Scholar
    • Export Citation
  • Marschner, R. J. 1950. Major Land Uses in the United States. United States Department of Agriculture Research Service, 252 pp.

  • Meinel, G., , M. Netzband, , V. Amann, , R. Stätter, , and G. Kritikos. 1996. Analysing an ATM-Scanner flight over the city of Dresden to identify urban sealing. Int. Arch. Photogramm. Remote Sens. 31:486492.

    • Search Google Scholar
    • Export Citation
  • Mesev, V. 2003. Remotely sensed cities: An introduction. Remotely Sensed Cities, V. Mesev, Ed., Taylor and Francis, 1–19.

  • National Academies 2006. Surface Temperature Reconstructions for the Last 2,000 Years. National Academies Press, 160 pp.

  • Netzband, M., , W. L. Stefanov, , and C. L. Redman. 2007. Remote sensing as a tool for urban planning and sustainability—An introduction. Applied Remote Sensing for Urban Planning, Governance and Sustainability, M. Netzband, W. L. Stefanov, and C. Redman, Eds., Springer, 1–23.

    • Search Google Scholar
    • Export Citation
  • Oke, T. R. 1987. Boundary Layer Climates. 2nd ed. Routledge, 464 pp.

  • Pomerantz, M., , H. Akbari, , and J. T. Harvey. 2000. Cooler reflective pavements give benefits beyond energy savings: Durability and illumination. Rep. No. LBNL-45370, Berkeley, CA, 14 pp.

    • Search Google Scholar
    • Export Citation
  • Quattrochi, D. A., and M. K. Ridd. 1994. Measurement and analysis of thermal energy responses from discrete urban surfaces using remote sensing data. Int. J. Remote Sens. 15:19912022.

    • Search Google Scholar
    • Export Citation
  • Reddy, S., , D. Truax, , E. Kolstad, , and C. O’Hara. 2004. Comparing spectral and object based approaches for classification and transportation feature extraction from high resolution multispectral imagery. Proc. ASPRS 2004 Annual Conf., Denver, CO, American Society for Photogrammetry and Remote Sensing. [Available online at http://www.ncrste.msstate.edu/archive/projects/msu/sunil_asprs04.pdf.].

    • Search Google Scholar
    • Export Citation
  • Ridd, M. K. 1995. Exploring a V-I-S (vegetation-impervious surface-soil) model for urban ecosystem analysis through remote sensing: Comparative anatomy for cities. Int. J. Remote Sens. 16:21652185.

    • Search Google Scholar
    • Export Citation
  • Roberts, S. M., , T. R. Oke, , J. A. Voogt, , C. S. B. Grimmond, , and A. Lemonsu. 2003. Energy storage in a European city center. Proc. Fifth Int. Conf. on Urban Climate, Lodz, Poland, University of Lodz, 241–244.

    • Search Google Scholar
    • Export Citation
  • Roessner, S., , K. Segl, , U. Heiden, , and H. Kaufmann. 2001. Automated differentiation of urban surfaces based on airborne hyperspectral imagery. IEEE Trans. Geosci. Remote Sens. 39:15251532.

    • Search Google Scholar
    • Export Citation
  • Rose, S. L., , H. Akbari, , and H. Taha. 2003. Characterizing the fabric of urban environment: A case study of metropolitan Houston, Texas. Lawrence Berkeley National Laboratory Rep. LBNL-51448, 67 pp.

    • Search Google Scholar
    • Export Citation
  • Schöpfer, E., and M. Möller. 2006. Comparing metropolitan areas—A transferable object-based image analysis approach. Agric. For. Meteor. 77:263293.

    • Search Google Scholar
    • Export Citation
  • Siegel, R., and J. R. Howell. 1981. Thermal Radiation Heat Transfer. 2nd ed. McGraw-Hill, 868 pp.

  • Silva, H. R., , R. Bhardwaj, , P. E. Phelan, , J. S. Golden, , and S. Grossman-Clarke. 2009. Development of a zero-dimensional mesoscale thermal model for urban climate. J. Appl. Meteor. Climatol. 48:657668.

    • Search Google Scholar
    • Export Citation
  • Stefanov, W. L., and M. Netzband. 2005. Assessment of ASTER land cover and MODIS NDVI data at multiple scales for ecological characterization of an arid urban center. Remote Sens. Environ. 99:3143.

    • Search Google Scholar
    • Export Citation
  • Stefanov, W. L., and A. J. Brazel. 2007. Challenges in characterizing and mitigating urban heat islands—A role for integrated approaches including remote sensing. Applied Remote Sensing for Urban Planning, Governance, and Sustainability, M. Netzband, W. L. Stefanov, and C. Redman, Eds., Springer, 117–135.

    • Search Google Scholar
    • Export Citation
  • Stefanov, W. L., , M. S. Ramsey, , and P. R. Christensen. 2001. Monitoring urban land cover change: An expert system approach to land cover classification of semiarid to arid urban centers. Remote Sens. Environ. 77:173185.

    • Search Google Scholar
    • Export Citation
  • Taha, H. 1999. Modifying a mesoscale meteorological model to better incorporate urban heat storage: A bulk-parameterization approach. J. Appl. Meteor. 38:466473.

    • Search Google Scholar
    • Export Citation
  • United Nations 2006. World Urbanization Prospects: The 2006 Revision Data Table and Highlights. Department of Economic and Social Affairs, 21 pp.

    • Search Google Scholar
    • Export Citation
  • Voogt, J. A., and T. R. Oke. 1997. Complete urban surface temperatures. J. Appl. Meteor. 36:11171132.

  • Wentz, E. A., , W. L. Stefanov, , C. Gries, , and D. Hope. 2006. Land use and land cover mapping from diverse sources for an arid urban environments. Comput. Environ. Urban Syst. 30:320346.

    • Search Google Scholar
    • Export Citation
  • Wharton, S. 1987. A spectral-knowledge-based approach for urban land-cover discrimination. IEEE Trans. Geosci. Remote Sens. 25:272282.

    • Search Google Scholar
    • Export Citation
  • Yuan, F., and M. E. Bauer. 2006. Mapping impervious surface area using high resolution imagery: A comparison of object-oriented classification to per-pixel classification. Proc. American Society of Photogrammetry and Remote Sensing Annual Conf., Reno, NV, ASPRS, CD-ROM, 8 pp.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Outline of the role of land cover in relationship to urban biocomplexity.

  • View in gallery

    ASTER thermal image of the Phoenix metropolitan area filtered for the top 25% of surface temperatures on 3 Oct 2003 2230 LT. Areas in red correlate predominantly to paved surfaces and exposed bedrock outcrops (Golden 2004).

  • View in gallery

    Daytime ASTER image of a portion of the Phoenix metropolitan area (left) processed using the SAVI and (right) to map surface albedo (Golden and Kaloush 2006).

  • View in gallery

    Quickbird image of the subject industrial park.

  • View in gallery

    Analysis results for HMA pavements within an industrial area of the city of Tempe.

  • View in gallery

    Parameterization of HMA paved surfaces by engineered function. Objects with gray variations are nonpavement structures and not part of the analysis.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 17 17 2
PDF Downloads 12 12 2

Enhanced Classifications of Engineered Paved Surfaces for Urban Systems Modeling

View More View Less
  • 1 School of Sustainability, and National Center of Excellence on SMART Innovations for Urban Climate and Energy, Arizona State University, Tempe, Arizona
  • | 2 School of Sustainability, Arizona State University, Tempe, Arizona
  • | 3 Image Science and Analysis Laboratory, NASA Johnson Space Center, Houston, Texas
© Get Permissions
Full access

Abstract

There is a greater need than ever for the ability to accurately model urban system impacts resulting around the planet. Rapid urbanization is transforming landscapes from vegetation to an engineered infrastructure and thus altering land cover and land use. These alterations impact urban and global climate change, energy demand, human health, and ecological service functions. This article presents an overview of a refined land-cover classification protocol that seeks to refine current land-cover classifications of engineered paved surfaces. This new approach provides those who model urban systems and engineer the environment as well as other scientists and policy makers an expanded understanding of how intervention to the system can most effectively be accomplished through enhanced modeling. An object-oriented analysis regime is presented for an industrial park utilizing commercial software in conjunction with multispectral and panchromatic Quickbird satellite imagery. A detailed examination of hot-mix asphalt paved surfaces was undertaken in relation to the materials’ engineered function such as various types of streets, parking, etc. The results were validated using a commercial raster graphics editor and data analysis software as well as on-site inspections. An overall accuracy of 95% was achieved.

Abstract

There is a greater need than ever for the ability to accurately model urban system impacts resulting around the planet. Rapid urbanization is transforming landscapes from vegetation to an engineered infrastructure and thus altering land cover and land use. These alterations impact urban and global climate change, energy demand, human health, and ecological service functions. This article presents an overview of a refined land-cover classification protocol that seeks to refine current land-cover classifications of engineered paved surfaces. This new approach provides those who model urban systems and engineer the environment as well as other scientists and policy makers an expanded understanding of how intervention to the system can most effectively be accomplished through enhanced modeling. An object-oriented analysis regime is presented for an industrial park utilizing commercial software in conjunction with multispectral and panchromatic Quickbird satellite imagery. A detailed examination of hot-mix asphalt paved surfaces was undertaken in relation to the materials’ engineered function such as various types of streets, parking, etc. The results were validated using a commercial raster graphics editor and data analysis software as well as on-site inspections. An overall accuracy of 95% was achieved.

Introduction

Our planet’s population increased from 3 billion in 1959 to 6 billion by 1999, a doubling that occurred over 40 yr. The United Nations latest projections (United Nations 2006) imply we will have 9 billion inhabitants by 2042, an increase of 50% that will require only 43 yr. At the same time and for the first time in history, the world is predominantly urban. More than half of the planet’s population lives in cities, up 30% from 50 yr ago. Urban areas are gaining an estimated 67 million people per year. The same report indicates that, by 2030, approximately 5 billion people are expected to live in urban areas—60% of the projected global population of 8.3 billion. Along with increasing population comes rapid land-use–land-cover (LULC) change and increased materials use. Rapid urbanization is transforming landscapes from vegetation to an engineered infrastructure that reduces evapotranspiration and increases thermal-storage capacity (Voogt and Oke 1997; Golden et al. 2005). This impact is often manifested in micro- and mesoscale modifications to the thermal properties of the surface and atmosphere and can result in rapid change in the urban climate compared to adjacent rural regions, known as the urban heat island (UHI) effect (Oke 1987). The importance of the study of surface materials and their system interactions with urban climate and environmental sustainability therefore will only continue to gain importance as urbanization continues. This article is intended to provide engineers, geographers, planners, climatologists, meteorologists, and environmental scientists greater knowledge of the contributions and impacts resulting from the urban climate–engineered materials interactions (Figure 1) and the resulting urban system (National Academies 2006; Jo et al. 2009).

Surface energy budgets

Voogt and Oke (Voogt and Oke 1997) identified that the surface is critical to urban climatologic understanding, as the surface is the principal location of sinks of heat, mass, and momentum. Interactions within the materials–climate system, however, are not well understood. The National Academies (National Academies 2006) released a report on radiative forcing of climate change, stating “the mechanisms involved in land-atmosphere interactions are not well understood, let alone represented in climate models,” and that “policies associated with land management practices also need to consider their inadvertent effects on climate.” They called for researchers to integrate climate variables when developing policies to control air pollution and manage land.

Because of the strong influence of the surface on the conditions in the overlying atmosphere, intentional modifications to surface characteristics through engineering become critical determinants to altering urban climates and their effects such as the UHI. Influencing the surface energy budget is the solar energy flux arriving at sea level, which is about 1100 W m−2, and 1353 W m−2 at the top of the atmosphere (Siegel and Howell 1981). The various components of a typical surface energy budget (W m−2) were presented by Roberts et al. (Roberts et al. 2003):
i1087-3562-13-5-1-e1
where Q* is net all-wave radiation, QF is the anthropogenic heat flux, QH is the sensible heat flux, QE is the latent heat flux, ΔQS is the net storage flux, and ΔQA is the net horizontal heat advection.

Land cover and energy budget

The role of urban heat storage and the surface energy budget has been historically weighted as a sum of the individual surface materials and incorporated in mesoscale modeling applications, such as the fifth-generation Pennsylvania State University–National Center for Atmospheric Research Mesoscale Model (MM5), Weather Research and Forecasting (WRF) Model and the Objective Hysteresis Model (OHM) by Grimmond et al. (Grimmond et al. 1991) and the more user-friendly zero-dimensional model for urban climate (Silva et al. 2009). A grid-type approach to evaluate the storage was introduced by Kerschgens and Kraus (Kerschgens and Kraus 1990) and refined by Taha (Taha 1999) where the contribution of the canopy layer fluxes to an area at the bottom of the boundary layer is the weighted sum of the contribution of fluxes from individual surfaces. The need to increase the resolution of LULC information, both spatially and temporally, is vital to realize functionally reliable meteorological modeling (Grossman-Clarke et al. 2005; Stefanov and Brazel 2007).

Burley (Burley 1961) described land cover as “the vegetational and artificial constructions covering the land surface.” However, in terms of meteorological modeling at different scales, researchers must consider additional issues such as soils, exposed bedrock, and water. This could include seasonally wet areas, marshes, intermittent rivers, man-made lakes, agricultural soils, soft- and hard-rock bedrock outcrops, or man-made surface mining areas, etc. Thus, a more inclusive definition is “biophysical materials found on the land.” In contrast, land use can be defined as “how the land is being used by human beings” (Jensen 2000).

In the 1940s an effort was made to begin to quantify land cover and land use. Major land-use mapping was conducted using aerial photographs. The most notable of this work was done by Marschner (Marschner 1950) who created state land-use maps at a scale of 1:5 000 000. With the advent of orbital satellite remote sensing, the United States Geological Survey attempted to standardize land-use and land-cover observations. Anderson et al. (Anderson et al. 1976) recognized that “there is no ideal classification of land use and land cover, and it is unlikely that one could ever be developed.” They did, however, provide an initial framework for a multilevel classification. Levels I and II were large regional spatial areas such as those at the national or state level. At the time, multispectral data were mostly derived from Landsat satellite imagery. Levels III and IV corresponded more closely with intrastate, county, or municipal levels. This resulted in a categorization for urban areas such that level I was classified as “urban or built-up land” with level II listed as

  • 11) residential;
  • 12) commercial and services;
  • 13) industrial (the focus of this research);
  • 14) transportation, communications, and utilities;
  • 15) industrial and commercial complexes;
  • 16) mixed urban or built-up land;
  • 17) other urban or built-up land.
Subcategorization of these subsets was derived for residential areas, which included criteria of capacity, type, and permanency of residence as the discriminating factors among classes:
  • 111) single-family units,
  • 112) multifamily units,
  • 113) group quarters,
  • 114) residential hotels,
  • 115) mobile home parks,
  • 116) transient lodgings,
  • 117) other.
Their work provided a significant contribution to classifying land cover and land use from remotely sensed data. However, greater refinement is necessary to adequately adapt these classifications for modeling climate and environmental conditions, as well as adaptation and mitigation strategies, at the local and regional scales. Specifically, various types of engineered materials and vegetation make up the urban fabric irrespective of the land use being residential, commercial, industrial mixed use, etc.

An in-depth review of the history and techniques of urban LULC classification using remotely sensed data is beyond the scope of this article, particularly as the topic is covered extensively elsewhere (Donnay et al. 2001; Mesev 2003; Netzband et al. 2007). Urban LULC analysis typically applies statistical per-pixel classification approaches of varying degrees of sophistication (e.g., Haack et al. 1987; Gong and Howarth 1990; Ridd 1995; Foody 2000; Wentz et al. 2006) or spectral analysis (e.g., Wharton 1987; Quattrochi and Ridd 1994; Meinel et al. 1996; Roessner et al. 2001; Herold et al. 2004) using passive orbital and airborne multispectral to hyperspectral data [i.e., astronaut photography, Landsat Multispectral Scanner (MSS)/Thematic Mapper (TM)/Enhanced Thematic Mapper Plus (ETM+), Quickbird, Moderate Resolution Imaging Spectroradiometer (MODIS), MODIS/Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) Airborne Simulator (MASTER), etc.; see Table 1 below for sensor acronym definitions]. Passive sensors measure reflectance and emittance of incident solar energy from the land surface, whereas active sensors are equipped with their own tunable and directional energy source (such as radar and lidar sensor systems). More recently, active sensor approaches to classifying the urban biophysical fabric have come to the forefront as radar and lidar data become more accessible to both researchers and municipalities (e.g., Dell’Acqua et al. 2003; Barnsley et al. 2003).

We provide several examples of the application of remotely sensed data to urban LULC analysis to highlight the importance of this information for calculation of surface energy budgets. A preliminary study was undertaken to quantify the urban fabric by examining four major land types through the use of aerial color orthophotography, the global biosphere emissions and interactions system (GLOBEIS) model data, and LULC information from the U.S. Geological Survey (USGS). The four types examined in the greater city of Houston, Texas, area included 1) commercial, 2) industrial, 3) educational, and 4) residential. Rose et al. (Rose et al. 2003) identified that vegetation covered about 39% of the area, roofs cover approximately 21%, and paved surfaces cover 29%. Akbari et al. (Akbari et al. 1999) and Akbari (Akbari 2001) examined the urban fabric of Sacramento, California, which revealed 39% of the area not occluded by the urban canopy (tree canopy) consisted of paved surfaces including roads, parking areas, and sidewalks. An evaluation of the entire metropolitan areas of Salt Lake City, Utah; Sacramento; and Chicago, Illinois; revealed the percentage of paved areas ranged from 30% to 39% as seen above the canopy and 36% to 45% viewed under the canopy layer. In residential areas, the paved surfaces were slightly lower at 29%–32% (Akbari et al. 1999; Rose et al. 2003).

A more sophisticated approach was developed by Stefanov et al. (Stefanov et al. 2001), who created an expert system using 30-m Landsat TM data to derive a land-cover classification for the semiarid Phoenix, Arizona, region. The expert system approach provides for the integration of remotely sensed data with other sources of georeferenced information such as land-use data, spatial texture, and digital elevation models (DEMs) to obtain greater classification accuracy. Logical decision rules are used with the various datasets to assign class values to each pixel. TM reflectance data acquired in 1998 [visible to shortwave infrared (VSWIR) bands plus a vegetation index “band”] were classified for land cover using a maximum likelihood decision rule. In addition, spatial texture of the TM data and a soil adjusted vegetation index (SAVI) was calculated. They developed an expert system to perform postclassification recoding of the initial land-cover classification using additional spatial datasets such as texture, land use, water rights, city boundaries, and Native American reservation boundaries. Pixels were reclassified using logical decision rules into 12 classes. The overall accuracy of this technique was 85%. Individual class user’s accuracy ranged from 73% to 99%.

However, the commercial/industrial materials class performed at a very poor 49% accuracy primarily because of the similarity of subpixel components with other classes. In further work for the eastern Phoenix metropolitan area using 15-m ASTER visible to near-infrared (VNIR) data acquired in 2000, Stefanov and Netzband (Stefanov and Netzband 2005) used a similar expert system that incorporated surface texture, vegetation index, and land-use information as ancillary data. A revised land-cover classification scheme more strictly focused on land-cover classes (relative to Stefanov et al. 2001) was also used. Notably, the commercial/industrial class used in Stefanov et al. (Stefanov et al. 2001) was disaggregated to asphalt, soil, vegetation, and reflective roof surface classes in Stefanov and Netzband (Stefanov and Netzband 2005) to take advantage of the higher spatial resolution of ASTER data. Use of the higher spatial resolution ASTER data, modified expert system, and revised classification scheme resulted in overall accuracy of 88%, with individual class user accuracy ranging from 81% (grass and shrubs) to 98% (water and xeric built classes).

Rationale for a more refined land-cover classification

Based on the results of the various efforts undertaken to this point, we undertook to identify a methodology to enhance the resolution of urban land cover based on engineered material input requirements for climatic and biocomplexity modeling schemes. For this effort, we focused on engineered materials, as prior research discussed within this paper indicates they represent the largest percentage of the urban fabric and considering their contribution to the urban heat island effect and related impacts.

As has been presented, multiple efforts have been undertaken to develop land-cover classification algorithms that are generally applicable to urban centers and to quantify the structure of the urban fabric. To refine land-cover contributions to climate and the environment, the National Center of Excellence on Urban Climate and Energy at Arizona State University (ASU) (see http://www.asusmart.com) has been collecting and analyzing multispectral daytime and nighttime ASTER imagery of global urban centers in collaboration with the joint National Aeronautics and Space Administration (NASA)–ASU “100 cities” program (additional information is available online at http://100cities.asu.edu/index.html). The ASTER currently provides the highest day/night thermal data resolution available from an orbiting platform. Unlike the ETM+ sensor on Landsat-7, which has one thermal channel (10.4–12.5 μm), ASTER has five channels covering 8.125 μm–11.65 μm. In addition to ASTER, other remotely sensed datasets from other satellite and airborne platforms are used in ongoing urban analysis at ASU (Table 1). Of particular note are the MODIS sensors on board both the NASA Terra and Aqua satellites. Terra’s orbit around the Earth is timed so that it passes from north to south across the equator in the morning, while Aqua passes south to north over the equator in the afternoon. This provides a revisit time over the entire Earth’s surface every 1–2 days, in contrast to single-platform sensors such as ASTER that can typically only acquire data over a given region of interest twice a month for either a day or night image.

The limitations of using multispectral remotely sensed data become apparent when researchers seek to correlate the various existing engineered surfaces to refined models at the local and mesoscales. As Figures 2 and 3 illustrate, the thermal bands of ASTER at 90 m per pixel resolution can support a hierarchal understanding of surface temperature variations at the regional or mesoscale for correlation to broad land-cover patterns. Figure 2 provides a visual representation of the top 25% of surface temperatures in the Phoenix, Arizona, metropolitan region. The image on the left of Figure 3 examines SAVI for the metropolitan Phoenix region while the right side of Figure 3 examines the albedo of the region. In both images, polygons are used to measure different urban morphologies in the region. Although useful, these resolutions do not provide for a high enough spatial resolution of land cover to permit researchers the ability to quantify land cover by engineered material type, which is necessary to account for climatic contributions and mitigation strategy modeling.

To address these limitations we propose a methodology of enhancing land-cover classification for the explicit utilization of mesoscale climate and biocomplexity modeling. The first step is to refine the understanding of the thermal and radiative properties of engineered pavements by laboratory and field validation, and the second step is to incorporate accurately classified land cover derived from very high-resolution commercially available visible to near-infrared multispectral information from 2.4 m per pixel Quickbird data using the eCognition software package.

Pavement variability

Simple default parameterization of residential, commercial, or industrial urban areas fails to accurately reflect the extensive array of engineered material types and their diurnal thermal behavior. This need for refined understanding of engineered materials in the urban region is gaining greater importance as local and regional governments are experimenting with newly developed types of engineered materials and new pavement surfaces to mitigate the UHI and achieve sustainable urban systems. When examining paved surface types in energy budgets, there are two dominant materials that need to be considered (Golden and Kaloush 2006). One of these is asphalt concrete (AC), which is a concrete (93% by weight) with an asphalt binder (7% by weight); this is also known as hot-mix asphalt (HMA). The other is Portland cement concrete (PCC), which is a concrete aggregate (56% by weight), sand (33%), and Portland cement binder (11% by weight). There is also the less-used chip sealing (CS) methodology, where asphalt is the binder and aggregate chips are spread on top. The aggregate provides strength while the binder acts as glue and provides stiffness. There are various composite designs that incorporate additives to these base materials. These include the utilization of a crumb rubber friction course of approximately 1 in. (2.54 cm) on top of a PCC base, primarily for reducing noise, as well as ultrathin whitetopping (UTW), where a 2–4-in. PCC layer is placed on top of an existing HMA base to provide mechanistic strength and performance. AC has been historically believed to be the major paving material for urban streets by a ratio of 9 to 1 (Pomerantz et al. 2000).

Each design has varied depths to support mechanistic properties (Table 2). Generally, HMA for parking lots has a depth of 4 in. (10.2 cm), and residential and arterial streets have depths of at least 6 in. (15.2 cm). Highway pavements typically have thicknesses of 12 in. PCC (30.5 cm), PCC sidewalks are 4 in. (10.2 cm), and runways are generally 21 in. PCC (53.3 cm). HMA is used for taxiways and aprons at airports and generally has depths up to 21 in. (53.3 cm) (Golden and Kaloush 2006). Each of these different surface types causes diurnal variability. As an example, Gui et al. (Gui et al. 2007b) modeled the diurnal benefits of utilizing an asphalt rubber–asphalt concrete (ARAC) pavement primarily for parking lots and streets, not for highways. The first evaluation was to determine what thermal benefits would be obtained during the daytime versus nighttime impacts, since ARAC has a higher volumetric specific heat capacity than conventional hot-mix asphalt. The 0.1905-m-thick modified ARAC had a cooling effect of ∼4°C at peak temperature and a 2°C higher temperature during the cooler early morning hours as compared to a conventional HMA.

An additional parameterization includes subsurface pavement structures including base courses. Although not part of this research paper, surface modeling can be easily achieved by following methodologies presented by Gui et al. (Gui et al. 2007a; Gui et al. 2007b).

Object-based land-cover classification system

To enhance the classification of land cover with increased knowledge of radiative and thermal characterization of surface materials, we undertook an object-based land-cover classification scheme. The hypothesis was that, by incorporating remote sensing data derived from satellite platforms with software that can analyze an image not by single pixels but through image objects and their mutual relations and attributes, we can derive a higher quality urban land-cover classification relative to single-pixel approaches. We examined a built-out representative one-square-mile industrial region located within the city of Tempe, a college town located adjacent to and just to the east of the city of Phoenix, Arizona. An industrial area was utilized for this initial analysis as the complexity of land cover is somewhat reduced through the absence of parks, schools, water bodies, and residential areas with different landscaping.

The remotely sensed data used were obtained by the Quickbird satellite on 5 July 2005. Quickbird is a high-resolution commercial Earth-observation satellite owned by DigitalGlobe and launched in 2001 as the first satellite in a constellation of three scheduled to be in orbit by 2008. Imagery is recorded in four spectral bands: blue (0.45–0.52 μm), green (0.52–0.60 μm), red (0.63–0.69 μm), and near infrared (0.76–0.90 μm) by a multispectral scanner with a spatial resolution of 2.4 m per pixel. At the same time, a panchromatic image is recorded within the wavelength of 0.44–0.90 μm with a spatial resolution of 0.6 m per pixel.

The Quickbird image used was a standard imagery product, which is a georeferenced datum. The corrections applied to the standard imagery are radiometrical, sensor, and geometrica. The image is projected to the Universal Transverse Mercator (UTM) system. The image has an average absolute geolocation accuracy of 23 m, a circular error (CE) of 90%, and a 14-m root-mean-square error (RMSE).

Both the multispectral and panchromatic bands of Quickbird were used as layers (Figure 4). Both layers were then analyzed through the use of the eCognition software package, which uses an object-oriented analysis approach. Scenes are segmented into homogeneous polygons (or image objects) at several different hierarchical levels and spatial scales by applying fuzzy classification algorithms to a given object’s spectral value, shape, texture, and boundary relationships with other neighboring objects defined by users (Baatz et al. 2004; Blaschke et al. 2004; Schöpfer and Möller 2006; Yuan and Bauer 2006). Various small objects on the ground of high-resolution images usually result in a motley classification outcome, which may cause misclassification and confusion when the researcher analyzes certain objects, such as shadows and water (Lewinski and Zaremski 2004). Using eCognition, we could classify images by entire objects instead of single pixel; this increases the accuracy of classification and ease of interpretation of results.

After obtaining the Quickbird data, a general object-based land-cover classification for an industrial area was developed building upon Anderson et al. (Anderson et al. 1976). In essence, level II data for industrial (category 13) would be subcategorized as presented in Table 3. For this initial and limited evaluation we focused solely on transportation-related surfaces. These pavement surfaces are categorized into four classes based on Federal Highway Administration Functional Classification Guidelines (Federal Highway Administration 2007). The guidelines define streets and highways by their connectivity and capacity. The hierarchy of the urban functional systems includes a principal arterial system, minor arterial streets system, collector street system, and local street system. According to the guidelines, the principal arterial system carries the major portion of trips entering and leaving the urban area, as well as the majority of through movements desiring to bypass the central city; the minor arterial street system interconnects with and augments the urban principal arterial system and provides service to trips of moderate length at a lower level of travel mobility than principal arterials; the collector street system provides both land access service and traffic circulation within residential neighborhoods, commercial and industrial areas; and the local street system comprises all facilities not on one of the higher systems (Federal Highway Administration 2007).

Multiresolution segmentation formation of objects was then utilized to extract objects using three criteria: shape, color, and scale parameter. Scale parameterization is the most important segmentation value that affects the size and output number of objects. The industrial area defined for evaluation is composed of large amounts of paved surfaces, primarily HMA for streets and parking. Some of the complexity involved in using multiresolution segmentation became apparent when adjacent pavements of different uses (surface parking versus street) could be extracted as one object and classified in the same class because of their similar spectral value. In addition, similar pavement surfaces with different albedo also increase the challenge of classification. Such is the case when environmental and radiative forces influence surface properties. For instance, surface reflectances of HMA pavements are influenced by traffic load, sky view, date of installation, etc. Figure 5 provides a visual representation of all HMA-classified areas within the subject industrial area of the city of Tempe. In Figure 5, most parking lots and roads are included except for several parking lots with a higher albedo. Each function has a separate and potentially unique thermal characteristic for model inputs.

To address the problems mentioned above and provide a greater accuracy, our research adopted a manual correction and adjustment scheme on the high-resolution image after segmentation. We referred to segmentation criteria values from prior works of transportation feature extractions (Reddy et al. 2004) and metropolitan area land-cover analysis (Schöpfer and Möller 2006), which utilize a higher weighting for color value. Furthermore, we utilized a higher scale parameter value at the same level in order to obtain a better segmentation result. Scale is heterogeneity tolerance within a segment. Color, smoothness, and compactness are all variables that optimize the segment’s spectral homogeneity and spatial complexity. Figure 6 is the final classification result generated by eCognition, and it provides a representation of HMA surfaces by function.

Accuracy assessment and field verification

The accuracy assessment was performed in Earth Resource Data Analysis System (ERDAS) Imagine, an advanced raster graphics editor and raster data analysis software. After obtaining classification results from eCognition, the classification information layer was converted into ERDAS Imagine. As samples for conducting accuracy assessment and field verification, 300 random points were generated (50 for each class). An on-site inspection was conducted on 27 February 2008. Through inputting the reference data collected we computed the overall accuracy as 95% with 0.9406 kappa statistics.

User accuracy for each class ranges from 90% to 100%, except for parking lots, which have an 84% accuracy. This represents only eight inaccurate classifications. However, all these eight points fall into the same cluster of covered storage buildings, which are similar in shape and albedo as that of covered surface parking.

Discussion and recommendation

The result of this new parameterization methodology provides researchers with the ability to more easily discern and classify polygons by their function, engineered structure, and boundary.

The increased availability of commercially distributed remote sensing products and software analysis tools provides researchers with greater access and ability to refine our knowledge of urban areas at a time of increased demand for innovative policies and designs to address urban and global climate change impacts. As presented above, there exist opportunities to provide policy makers with a greater understanding of how future land-cover and land-use alterations (status quo or deliberate mitigation strategies) will impact the urban climate system and interact with the more extensive biocomplexity system.

The use of an integrated framework that incorporates traditional engineering functions of pavement analysis in conjunction with emerging remote sensing products and analysis tools has been found to be beneficial to those examining complex systems.

The approach of using high-resolution satellite images to provide clear boundaries and shapes of features augmented by segmenting images was an effective method for undertaking manual correction and visual interpretations. Lessons learned included 1) generating fewer polygons without losing too much information and 2) forming polygons that closely follow natural feature boundaries. The first consideration reduces the complexity and time to process, which can be extensive for larger areas of examination. The second consideration forms more completed polygons along with the natural feature shape instead of fragile pieces. This allows for a linear road, a square parking lot, or a rectangular building to be selected by “one click” in the processing software.

Scale parameterization is the most important input that affects the size and numbers of polygons. According to our observation, a scale parameter above 200 is better to form relatively completed polygons for most buildings, and 500 is more effective for roads. When the scale parameter is above 500, the accuracy is reduced because of loss of detailed information on the Quickbird image.

Color is the variable that considers a feature’s spectral homogeneity during the segmentation. A higher weight of color variable makes polygons follow the feature boundary. On the contrary, a higher weight of shape homogeneity leads to amorphous polygons (Benz et al. 2004). In our research, we found that a weight above 70% in color for extracting features by their natural boundaries is necessary.

Future efforts will combine a more extensive analysis of coupled remote sensing and laboratory quantifications of engineered pavements and building materials that would be found in residential and commercial areas as well as quantify land-cover vegetation and water bodies.

With these analyses, the research team will conduct evaluations of how different land covers in multiple urban regions interact with human health vulnerability, regional climate, electricity consumption, etc. Based on these findings we anticipate that local and regional governments can more effectively develop risk analyses and cost-effective adaptation and mitigation strategies for urban heat islands and other biocomplexity systems such as the nexus of urban climate and electricity reliability.

Acknowledgments

This work was supported in part by the U.S. Environmental Protection Agency Urban Heat Island Initiative (Number EP06H000497), the National Center for Environmental Health at the U.S. Centers for Disease Control and Prevention (Contract 30-07184-03 CDC/Task Order 0078), and the National Center of Excellence on SMART Innovations for Urban Climate and Energy (www.asuSMART.org).

REFERENCES

  • Akbari, H., , L. Rose, , and H. Taha. 1999. Characterizing the fabric of the urban environment: A case study of Sacramento, California. Lawrence Berkeley National Laboratory, U.S. Department of Energy, 48 pp.

    • Search Google Scholar
    • Export Citation
  • Akbari, H., , M. Pomerantz, , and H. Taha. 2001. Cool surfaces and shade trees to reduce energy use and improve air quality in urban areas. Sol. Energy 70:295310.

    • Search Google Scholar
    • Export Citation
  • Anderson, J., , E. Hardy, , J. Roach, , and R. Witmer. 1976. A land use and land cover classification system for use with remote sensor data. Geological Survey Professional Paper 964, United States Department of the Interior–Geological Survey, United States Government Printing Office, 24 pp.

    • Search Google Scholar
    • Export Citation
  • Baatz, M. Coauthors 2004. eCognition User Guide 4. Definiens Imaging. Munchen, Germany. [Available online at http://www.gis.unbc.ca/help/software/ecognition4/ELuserguide.pdf.].

    • Search Google Scholar
    • Export Citation
  • Barnsley, M. J., , A. M. Steel, , and S. L. Barr. 2003. Determining urban land use through an analysis of the spatial composition of buildings identified in LIDAR and multispectral image data. Remotely Sensed Cities, V. Mesev, Ed., Taylor and Francis, 83–108.

    • Search Google Scholar
    • Export Citation
  • Benz, U. C., , P. Hofmann, , G. Willhauck, , I. Lingenfelder, , and M. Heynen. 2004. Multiresolution, object-oriented fuzzy analysis of remote sensing data for GIS-ready information. ISPRS J. Photogramm. Remote Sens. 58:239258.

    • Search Google Scholar
    • Export Citation
  • Blaschke, T., , C. Burnett, , and A. Pekkarinen. 2004. Image segmentation methods for object-based analysis and classification. Remote Sensing Image Analysis: Including the Spatial Domain, F. De Meer and S. de Jong, Eds., Kluver, 211–236.

    • Search Google Scholar
    • Export Citation
  • Burley, T. M. 1961. Land use or land utilization? Prof. Geographer 13:1820.

  • Dell’Acqua, F., , P. Gamba, , and G. Lisini. 2003. Improvements to urban area characterization using multitemporal and multiangle SAR images. IEEE Trans. Geosci. Remote Sens. 41:19962004.

    • Search Google Scholar
    • Export Citation
  • Donnay, J-P., , M. J. Barnsley, , and P. A. Longley. 2001. Remote sensing and urban analysis. Remote Sensing and Urban Analysis, J.-P. Donnay, M. J. Barnsley, and P. A. Longley, Eds., Taylor and Francis, 3–18.

    • Search Google Scholar
    • Export Citation
  • Federal Highway Administration cited. 2007. FHWA functional classification guidelines. (Created 1989.) [Available online at http://tpd.azdot.gov/gis/fclass/fc_fhwa_sect_2_1.php.].

    • Search Google Scholar
    • Export Citation
  • Foody, G. M. 2000. Estimation of sub-pixel land cover composition in the presence of untrained classes. Comput. Geosci. 26:469478.

  • Golden, J. S. 2004. The built environment induced urban heat island effect in rapidly urbanizing arid regions—A sustainable urban engineering complexity. Environ. Sci.: J. Integr. Environ. Res. 1:321349.

    • Search Google Scholar
    • Export Citation
  • Golden, J. S., and K. Kaloush. 2006. Mesoscale and microscale evaluations of surface pavement impacts to the urban heat island effects. Int. J. Pavement Eng. 7:3752.

    • Search Google Scholar
    • Export Citation
  • Golden, J. S., , P. Guthrie, , K. Kaloush, , and R. Britter. 2005. The summertime urban heat island hysteresis lag complexity: Applying thermodynamics, urban engineering and sustainability research. Sustainable Eng.—J. Roy. Inst. Civ. Eng. 158:197210.

    • Search Google Scholar
    • Export Citation
  • Gong, P., and P. J. Howarth. 1990. The use of structural information for improving land-cover classification accuracies at the rural-urban fringe. Photogramm. Eng. Remote Sens. 56:6773.

    • Search Google Scholar
    • Export Citation
  • Grimmond, C. S. B., , H. A. Cleugh, , and T. R. Oke. 1991. An objective urban heat storage model and its comparison with other schemes. Atmos. Environ. 25B:311326.

    • Search Google Scholar
    • Export Citation
  • Grossman-Clarke, S., , J. A. Zehnder, , W. L. Stefanov, , Y. Liu, , and M. A. Zoldak. 2005. Urban modifications in a mesoscale meteorological model and the effects on near-surface variables in an arid metropolitan region. J. Appl. Meteor. 44:12811297.

    • Search Google Scholar
    • Export Citation
  • Gui, J., , J. Carlson, , P. Phelan, , K. Kaloush, , and J. S. Golden. 2007a. Impact of pavement thickness on surface diurnal temperatures. J. Green Build. 2:121130.

    • Search Google Scholar
    • Export Citation
  • Gui, J., , P. Phelan, , K. Kaloush, , and J. Golden. 2007b. Impact of pavement thermo-physical properties on surface temperatures. J. Mater. Civ. Eng. 19:683690.

    • Search Google Scholar
    • Export Citation
  • Haack, B., , N. Bryant, , and S. Adams. 1987. An assessment of Landsat MSS and TM data for urban and near-urban land-cover digital classification. Remote Sens. Environ. 21:201213.

    • Search Google Scholar
    • Export Citation
  • Herold, M., , D. A. Roberts, , M. E. Gardner, , and P. E. Dennison. 2004. Spectrometry for urban remote sensing—Development and analysis of a spectral library from 350 to 2400 nm. Remote Sens. Environ. 91:304319.

    • Search Google Scholar
    • Export Citation
  • Jensen, J. R. 2000. Remote Sensing of the Environment: An Earth Resource Perspective. Prentice-Hall, 544 pp.

  • Jo, J., , J. S. Golden, , and S. W. Shin. 2008. Incorporating built environment factors into climate change mitigation strategies for Seoul, South Korea: A sustainable urban systems framework. Habitat Int. 33:267275.

    • Search Google Scholar
    • Export Citation
  • Kerschgens, M. J., and H. Kraus. 1990. On the energetics of the urban canopy layer. Atmos. Environ. 24B:321328.

  • Lewinski, S., and K. Zaremski. 2004. Examples of object-oriented classification performed on high-resolution satellite images. Misc. Geogr. 11:349358.

    • Search Google Scholar
    • Export Citation
  • Marschner, R. J. 1950. Major Land Uses in the United States. United States Department of Agriculture Research Service, 252 pp.

  • Meinel, G., , M. Netzband, , V. Amann, , R. Stätter, , and G. Kritikos. 1996. Analysing an ATM-Scanner flight over the city of Dresden to identify urban sealing. Int. Arch. Photogramm. Remote Sens. 31:486492.

    • Search Google Scholar
    • Export Citation
  • Mesev, V. 2003. Remotely sensed cities: An introduction. Remotely Sensed Cities, V. Mesev, Ed., Taylor and Francis, 1–19.

  • National Academies 2006. Surface Temperature Reconstructions for the Last 2,000 Years. National Academies Press, 160 pp.

  • Netzband, M., , W. L. Stefanov, , and C. L. Redman. 2007. Remote sensing as a tool for urban planning and sustainability—An introduction. Applied Remote Sensing for Urban Planning, Governance and Sustainability, M. Netzband, W. L. Stefanov, and C. Redman, Eds., Springer, 1–23.

    • Search Google Scholar
    • Export Citation
  • Oke, T. R. 1987. Boundary Layer Climates. 2nd ed. Routledge, 464 pp.

  • Pomerantz, M., , H. Akbari, , and J. T. Harvey. 2000. Cooler reflective pavements give benefits beyond energy savings: Durability and illumination. Rep. No. LBNL-45370, Berkeley, CA, 14 pp.

    • Search Google Scholar
    • Export Citation
  • Quattrochi, D. A., and M. K. Ridd. 1994. Measurement and analysis of thermal energy responses from discrete urban surfaces using remote sensing data. Int. J. Remote Sens. 15:19912022.

    • Search Google Scholar
    • Export Citation
  • Reddy, S., , D. Truax, , E. Kolstad, , and C. O’Hara. 2004. Comparing spectral and object based approaches for classification and transportation feature extraction from high resolution multispectral imagery. Proc. ASPRS 2004 Annual Conf., Denver, CO, American Society for Photogrammetry and Remote Sensing. [Available online at http://www.ncrste.msstate.edu/archive/projects/msu/sunil_asprs04.pdf.].

    • Search Google Scholar
    • Export Citation
  • Ridd, M. K. 1995. Exploring a V-I-S (vegetation-impervious surface-soil) model for urban ecosystem analysis through remote sensing: Comparative anatomy for cities. Int. J. Remote Sens. 16:21652185.

    • Search Google Scholar
    • Export Citation
  • Roberts, S. M., , T. R. Oke, , J. A. Voogt, , C. S. B. Grimmond, , and A. Lemonsu. 2003. Energy storage in a European city center. Proc. Fifth Int. Conf. on Urban Climate, Lodz, Poland, University of Lodz, 241–244.

    • Search Google Scholar
    • Export Citation
  • Roessner, S., , K. Segl, , U. Heiden, , and H. Kaufmann. 2001. Automated differentiation of urban surfaces based on airborne hyperspectral imagery. IEEE Trans. Geosci. Remote Sens. 39:15251532.

    • Search Google Scholar
    • Export Citation
  • Rose, S. L., , H. Akbari, , and H. Taha. 2003. Characterizing the fabric of urban environment: A case study of metropolitan Houston, Texas. Lawrence Berkeley National Laboratory Rep. LBNL-51448, 67 pp.

    • Search Google Scholar
    • Export Citation
  • Schöpfer, E., and M. Möller. 2006. Comparing metropolitan areas—A transferable object-based image analysis approach. Agric. For. Meteor. 77:263293.

    • Search Google Scholar
    • Export Citation
  • Siegel, R., and J. R. Howell. 1981. Thermal Radiation Heat Transfer. 2nd ed. McGraw-Hill, 868 pp.

  • Silva, H. R., , R. Bhardwaj, , P. E. Phelan, , J. S. Golden, , and S. Grossman-Clarke. 2009. Development of a zero-dimensional mesoscale thermal model for urban climate. J. Appl. Meteor. Climatol. 48:657668.

    • Search Google Scholar
    • Export Citation
  • Stefanov, W. L., and M. Netzband. 2005. Assessment of ASTER land cover and MODIS NDVI data at multiple scales for ecological characterization of an arid urban center. Remote Sens. Environ. 99:3143.

    • Search Google Scholar
    • Export Citation
  • Stefanov, W. L., and A. J. Brazel. 2007. Challenges in characterizing and mitigating urban heat islands—A role for integrated approaches including remote sensing. Applied Remote Sensing for Urban Planning, Governance, and Sustainability, M. Netzband, W. L. Stefanov, and C. Redman, Eds., Springer, 117–135.

    • Search Google Scholar
    • Export Citation
  • Stefanov, W. L., , M. S. Ramsey, , and P. R. Christensen. 2001. Monitoring urban land cover change: An expert system approach to land cover classification of semiarid to arid urban centers. Remote Sens. Environ. 77:173185.

    • Search Google Scholar
    • Export Citation
  • Taha, H. 1999. Modifying a mesoscale meteorological model to better incorporate urban heat storage: A bulk-parameterization approach. J. Appl. Meteor. 38:466473.

    • Search Google Scholar
    • Export Citation
  • United Nations 2006. World Urbanization Prospects: The 2006 Revision Data Table and Highlights. Department of Economic and Social Affairs, 21 pp.

    • Search Google Scholar
    • Export Citation
  • Voogt, J. A., and T. R. Oke. 1997. Complete urban surface temperatures. J. Appl. Meteor. 36:11171132.

  • Wentz, E. A., , W. L. Stefanov, , C. Gries, , and D. Hope. 2006. Land use and land cover mapping from diverse sources for an arid urban environments. Comput. Environ. Urban Syst. 30:320346.

    • Search Google Scholar
    • Export Citation
  • Wharton, S. 1987. A spectral-knowledge-based approach for urban land-cover discrimination. IEEE Trans. Geosci. Remote Sens. 25:272282.

    • Search Google Scholar
    • Export Citation
  • Yuan, F., and M. E. Bauer. 2006. Mapping impervious surface area using high resolution imagery: A comparison of object-oriented classification to per-pixel classification. Proc. American Society of Photogrammetry and Remote Sensing Annual Conf., Reno, NV, ASPRS, CD-ROM, 8 pp.

    • Search Google Scholar
    • Export Citation

Figure 1.
Figure 1.

Outline of the role of land cover in relationship to urban biocomplexity.

Citation: Earth Interactions 13, 5; 10.1175/2009EI274.1

Figure 2.
Figure 2.

ASTER thermal image of the Phoenix metropolitan area filtered for the top 25% of surface temperatures on 3 Oct 2003 2230 LT. Areas in red correlate predominantly to paved surfaces and exposed bedrock outcrops (Golden 2004).

Citation: Earth Interactions 13, 5; 10.1175/2009EI274.1

Figure 3.
Figure 3.

Daytime ASTER image of a portion of the Phoenix metropolitan area (left) processed using the SAVI and (right) to map surface albedo (Golden and Kaloush 2006).

Citation: Earth Interactions 13, 5; 10.1175/2009EI274.1

Figure 4.
Figure 4.

Quickbird image of the subject industrial park.

Citation: Earth Interactions 13, 5; 10.1175/2009EI274.1

Figure 5.
Figure 5.

Analysis results for HMA pavements within an industrial area of the city of Tempe.

Citation: Earth Interactions 13, 5; 10.1175/2009EI274.1

Figure 6.
Figure 6.

Parameterization of HMA paved surfaces by engineered function. Objects with gray variations are nonpavement structures and not part of the analysis.

Citation: Earth Interactions 13, 5; 10.1175/2009EI274.1

Table 1.

Remotely sensed datasets available for land-cover classification. Thermal or midinfrared wavelengths (TIR), vegetation indexes (VI), and surface temperature (ST).

Table 1.
Table 2.

Thermal variability examples of selected paved surfaces. (Sources: Golden and Kaloush 2006; Gui et al. 2007a; Gui et al. 2007b)

Table 2.
Table 3.

Classification results for HMA paved surfaces within an industrial area. The summation of all asphalt surfaces located within the research area and the relative percentage of HMA pavements within the industrial area by class.

Table 3.

* Corresponding author address: Jay Golden, School of Sustainability, Arizona State University, P.O. Box 875502, Tempe, AZ 85287-5502. jay.golden@asu.edu

Save