Local-scale fluxes of carbon dioxide in urban environments: methodological challenges and results from Chicago

https://doi.org/10.1016/S0269-7491(01)00256-1Get rights and content

Abstract

Much attention is being directed to the measurement and modeling of surface–atmosphere exchanges of CO2 for different surface types. However, as yet, few measurements have been conducted in cities, even though these environments are widely acknowledged to be major sources of anthropogenic CO2. This paper highlights some of the challenges facing micrometeorologists attempting to use eddy covariance techniques to directly monitor CO2 fluxes in urban environments, focusing on the inherent variability within and between urban areas, and the importance of scale and the appropriate height of measurements. Results from a very short-term study of CO2 fluxes, undertaken in Chicago, Illinois in the summer of 1995, are presented. Mid-afternoon minimum CO2 concentrations and negative fluxes are attributed to the strength of biospheric photosynthesis and strong mixing of local anthropogenic sources in a deep mixed layer. Poor night-time atmospheric mixing, lower mixed layer depths, biospheric respiration, and continued emissions from mobile and fixed anthropogenic sources, account for the night-time maxima in CO2 concentrations. The need for more, longer-term, continuous eddy covariance measurements is stressed.

Introduction

Urban areas represent a location where a large and ever increasing number of people live and where a disproportionate share of natural resources, including fossil fuels, are used. Attempts to quantify the role of urban areas on the global carbon budget have focused largely on inventories of emissions, from estimates of fossil fuel consumption, cement production, etc. (e.g. Mensink et al., 2000) and the amount of carbon sequestered in urban vegetation based on biomass estimates (Nowak, 1994a, Jo & McPherson, 1995), or short-term studies of CO2 concentrations, documenting spatial patterns across cities or at single sites through time (see examples in Table 1). While these studies have documented that CO2 concentrations are greater in urban environments, and isotopic analyses (Nakazawa et al., 1997, Kuc & Zimnoch, 1998) have attributed this difference to anthropogenic sources, none of the studies have documented the actual fluxes of CO2 and their diffusive characteristics in urban environments, essential for assessing the potential impact on climate and biosphere at all scales (Dabberdt et al., 1993).

Surface–atmosphere exchanges of CO2 can be measured directly using micrometeorological techniques, notably eddy covariance equipment mounted on tall towers. This approach has been employed for other ecosystems, notably grasslands, forests and wetlands, as part of the global FLUXNET program (Baldocchi et al., 2001a, Baldocchi et al., 2001b). Based on such measurements, important data are emerging on the role of these different ecosystems, spatial and temporal (daily, seasonal, and annual) variability and controls. However, enormous challenges face micrometeorologists trying to make meaningful flux observations using such technologies in urban environments. The spatial variability of surface cover and roughness is extreme, presenting special challenges to those wanting to make representative measurements, both in terms of siting equipment to appropriately measure sources and sinks, and subsequently when trying to generalize results to larger areas.

The purpose of this paper is two-fold: first, to highlight some of the key issues that need to be considered when making direct (micrometeorological) flux measurements of CO2 in urban environments, drawing on results and experience from measuring other fluxes, notably the latent heat flux (evapotranspiration) in cities; and second, to present select results from a short-term flux measurement campaign conducted in Chicago, Illinois in the summer of 1995. These are amongst the first measurements of CO2 fluxes made in urban environments, though it is important to stress that new initiatives to undertake long-term, continuous measurements in cities are now underway; see, for example, the Baltimore NSF-funded Urban Long Term Ecological research site.

Whether considered in terms of roughness (the size, shape and separation of buildings and vegetation) or surface cover (the spatial arrangement and the range of radiative, thermal and moisture properties), the broad category “urban” and the land uses within (commercial, downtown, industrial, suburban, etc.) commonly incorporate a wider range of surface characteristics than forests, agricultural areas, or wetlands. Observations from many carefully selected sites with contrasting surface cover, energy use, and traffic regimes will be needed to characterize CO2 fluxes in cities. In the context of CO2 fluxes, of particular relevance are variations in vegetation cover and photosynthetic activity, and emissions from fixed (industrial, commercial, institutional) and mobile (traffic) sources. Grimmond and Oke (1999b) have documented that within urban areas evapotranspiration rates can be significant but are highly variable (Fig. 1). The area vegetated, and even more so the area irrigated, exerts an important control on turbulent heat partitioning (see results summarized on the lower right of Fig. 1). Given the importance of irrigation for evapotranspiration, the area irrigated also would be expected to be important in controlling photosynthesis, and thus carbon uptake.

It is widely accepted that the understanding of urban climates, and their observation and modeling, is critically tied to notions of scale (spatial and temporal) and boundary layer development. For urban areas, three spatial scales (micro-, local-, and meso-) are commonly recognized (based on Oke, 1984), and provide a basis for appropriately siting equipment and generalizing results. At the micro-scale (101–102 m), important spatial differences in processes occur in response to variability in building/canyon dimensions and orientations and proximity to localized CO2 emissions (e.g. individual roads). At the local-scale (102–104 m), processes represent the integrated response of an array of buildings, vegetation, and paved surfaces. At this scale, spatial variability across a city reflects different neighborhoods, with various combinations of built and vegetated cover and morphometry. At the meso-scale (104–105 m), the city is considered in its entirety, and differentiated from its surroundings, areas of forest, agriculture, etc. Of those studies of CO2 concentrations conducted in urban environments to date (Table 1), virtually all, with the notable exceptions of Kuc, 1991, Nakazawa et al., 1997, Reid & Steyn, 1997, Kuc & Zimnoch, 1998, focus on the micro-scale, considering processes and patterns in the urban canyon (below building height). Inadequate attention has as yet focused on how micro-scale results can be extrapolated to larger scales and their implications for documenting the effect of urban areas regionally.

A common approach to documenting urban effects is to conduct simultaneous urban–rural measurements, with differences attributed to the effects of urbanization. Oke and Grimmond (2000), in a study of surface energy balance fluxes, draw on Lowry (1977) to show that unless great care is taken the approach is flawed, as proximal rural sites are themselves subjected to anthropogenic influences and may also be affected by advection of urban influences. In the context of CO2 fluxes and concentrations these influences may be significant; agriculture, for example, has a major effects on CO2 concentrations (e.g. Berry and Colls, 1990a).

Advances in instrumentation (notably eddy covariance technology) have meant that representative flux data can be collected from urban areas, provided careful attention is paid to the siting and operation of equipment (Oke et al., 1989, Grimmond & Oke, 1999a, Roth, 2000). Instruments must be mounted at a height at least twice the mean height of the roughness elements (buildings and trees) to ensure that the instruments are above the influence of individual roughness elements and the measurements represent an integrated response at the local-scale (Grimmond & Oke, 1999a, Kastner-Klein et al., 2000, Rotach, 2000). Vertical profiles of CO2 concentrations within the urban canopy are needed to account for storage changes between the surface and the flux measurement level. Fetch in dominant upwind directions (∼1–2 km) must be fairly uniform (similar patterns of buildings, roads, vegetation, etc.) so that controls related to particular surface covers and emissions sources can be identified, and effects of advection are minimal, so that networks of stations are not needed to document horizontal gradients.

Although closed path sensors have been used in many studies, open path sensors have many advantages and also allow continuous measurements of the flux of CO2. Closed path sensors require air to be sucked (pumped) to the ground for analysis with appropriate gas analyzers, and thus lag corrections (to account for the time lags between the measurements of the vertical velocity on the tower by a sonic anemometer and the gas concentration by the analyzer often at the ground) need to be made and pumps and flow controllers need to be maintained. Temporal lags may be greater in urban environments given the height needed to obtain representative measurements (although in mature forests the same issues are encountered, e.g. Schmid et al., 2000). As in other environments, issues related to the infilling of data, for periods with low friction velocity, will need to be addressed. If topographic variations occur in the vicinity of the site (common in many cities), the mean vertical wind velocity (w) will be difficult to determine because of drainage flows. On an annual basis the number of hours with these types of conditions may become large, resulting in a large systematic error (Lee, 1998, Paw et al., 1998, Finnigan, 1999). All these issues are receiving significant attention in micrometeorology today (e.g. Baldocchi et al., 2001a,b)

One of the few advantages of working in an urban environment is that detailed information on the surface (size, shape and spacing of roughness elements; fractional cover of different surface types—greenspace, roofs, impervious, etc.; emissions inventories of major CO2 sources) often is available in spatially referenced databases (Geographic Information Systems). Such databases can be sampled, by overlying meteorological source area (footprint) models, for example, Schmid's (1997) FSAM—flux source area model, to describe key sources and sinks of CO2. A general example of such an application, although not for CO2, is presented in Grimmond and Souch (1994).

Section snippets

Material and methods

Measurements of surface–atmosphere exchanges of CO2 were conducted in a northwest suburb of Chicago, Illinois (41° 57′ N 87° 48′ W) in the summer of 1995. Here we describe the site, equipment used, and details of the post-processing of the data.

Results and discussion

Fig. 3a presents average concentrations of CO2 for 13 days in the summer of 1995; plots for the individual days are shown in detail in Fig. 4. Rates of evapotranspiration (latent heat flux) for this period, measured using the krypton hygrometer, are shown in Fig. 1. In Chicago, the mean CO2 concentration for these days was 384 ppmv; this compares with background Mauna Loa values of 361.60 ppmv (Keeling and Whorf, 2000). Although variations between days are evident, a marked and distinct diurnal

Conclusion

At the end of the twentieth century, approximately half of the world's population, over three billion people, lived in urban areas. By 2025, the United Nations (cited in Uitto and Biswas, 2000) predicts that this number will double, and the proportion of the global population who are urban residents will rise to two-thirds. Urban areas are important sources of CO2, and locations where enhanced concentrations are amongst the most pronounced.

As yet, few measurements of concentrations or more

Acknowledgements

The assistance of the people who provided permission to use sites, and the many who aided with the fieldwork and data analysis are greatly appreciated. In particular, we would like to thank Emily Freeman, Kevin Sayers, Neal Schroeder and Mark Hubble. Dr. Tilden Myers provided invaluable advice on instrumentation and software. John Chin provided the 2000 Mauna Loa data. Funding was provided by USDA Forest Service co-operative research grants No. 23-526 and No. 23-546. This paper was presented at

References (52)

  • D. Baldocchi et al.

    On measuring net ecosystem carbon exchange over tall vegetation on complex terrain

    Boundary Layer Meteorology

    (2001)
  • D. Berry et al.

    Atmospheric carbon dioxide and sulphur dioxide on an urban/rural transect—I. Continuous measurements at the transect ends

    Atmospheric Environment

    (1990)
  • D. Berry et al.

    Atmospheric carbon dioxide and sulphur dioxide on an urban/rural transect—II. Measurements along the transect

    Atmospheric Environment

    (1990)
  • S.A. Changnon et al.

    Impacts and responses to the 1995 Heat Wavea call to action

    Bulletin of the American Meteorological Society

    (1996)
  • The Heat is On: America's CO2 Polluters (Research Report)

    (1992)
  • J.F. Clarke et al.

    An evaluation of CO2 measurements as an indicator of air pollution

    Journal Air Pollution Control Association

    (1966)
  • W.F. Dabberdt et al.

    Atmosphere-surface exchange measurements

    Science

    (1993)
  • Day, T.A., Mueller, E.C., Xiong, F.S., Balling, R.C., 2000. Temporal patterns in near surface CO2 concentrations in the...
  • B. Ghauri et al.

    An assessment of air-quality in Karachi, Pakistan

    Environmental Monitoring and Assessment

    (1994)
  • C.S.B. Grimmond et al.

    Surface description for urban climate studiesa GIS based methodology

    Geocarto International

    (1994)
  • C.S.B. Grimmond et al.

    Aerodynamic properties of urban areas derived from analysis of surface form

    Journal of Applied Meteorology

    (1999)
  • C.S.B. Grimmond et al.

    Evapotranspiration rates in urban areas

    In: Impacts of Urban Growth on Surface Water and Groundwater Quality, International Association of Hydrological Sciences

    (1999)
  • Grimmond, C.S.B., Oke, T.R., 2000. Heat fluxes and stability in North American cities. In: American Meteorological...
  • C.S.B. Grimmond et al.

    Surface aerodynamic characteristics of urban areasanemometric analyses

    Boundary Layer Meteorology

    (1998)
  • B.C. Halter et al.

    On the variability of atmospheric carbon dioxide concentration at Barrow, Alaska 1973–1979

    Tellus

    (1981)
  • C.D. Idso et al.

    The urban CO2 dome of Phoenix, Arizona

    Physical Geography

    (1998)
  • Cited by (239)

    • COVID-19 pandemic underscores role of green space in urban carbon dynamics

      2023, Science of the Total Environment
      Citation Excerpt :

      Overall, these studies point toward the potential of reduction in traffic emissions as an effective strategy in mitigating climate change (Liu et al., 2020; Nicolini et al., 2022). Generally speaking, the net carbon dioxide exchange between the land surface and the atmosphere in urban settings (NUE) does tend to be dominated by fossil fuel combustion (C) from vehicles, industry, and individual households rather than the biological processes of photosynthesis (P) and respiration (R) (Crawford and Christen, 2015; Grimmond et al., 2002). Therefore, the focus on emissions is logical.

    View all citing articles on Scopus
    View full text