Researchers: Marina Alberti and Lucy R. Hutyra 2012

Terrestrial carbon stocks across a gradient of urbanization a study of the Seattle, WA region
Problem Statement
  • Urbanizing regions are major determinants of global and continental scale changes in carbon budgets through land transformation, modification of related biogeochemical processes, and concentration of fossil fuel combustion activities (Pataki et al. 2007).
  • Recent studies of biogeochemistry in urbanizing regions provide evidence of the complex mechanisms by which urban activities affect C fluxes and stocks (Pataki et al. 2007; Churkina, Brown, and Keoleian 2010; Pickett et al. 2011).
  • There are few empirical data on the underlying mechanisms linking urban patterns and carbon budgets to systematically evaluate how alternative patterns of urban development interact with ecosystem processes across a gradient of urbanization.
Overarching Research Questions

1. What factors control changes in carbon stocks and fluxes along a gradient of urbanization (Box 1)?
2. What are the tradeoffs between stocks and fluxes associated with patterns of urbanization?
3. What are the uncertainties, lags, legacies, and feedbacks associated with urban land use and infrastructure decisions on carbon fluxes and stocks?
4. How will the interactions between urbanization patterns and carbon processes evolve under future scenarios?

Table 1
Figure 1
Figure 2
Figure 3
Figure 4
Figure 4
figure 5
Figure 5
figure 6
Figure 6
Figure 7
Figure 7
Conceptual Framework and Hypotheses

Building on mechanisms established in the literature that link urban patterns and the C cycle (see Table 1), we articulate a framework (Fig 1) and set of testable hypotheses on how these mechanisms vary across a hypothetical gradient of urbanization (Figure 2).

  • Hypothesis 1. Variability in carbon stocks and fluxes across gradients of urbanization is controlled by complex interactions between land cover, emissions, organic inputs, temperature, and N fertilization.
  • Hypothesis 2. Carbon fluxes vary across an urban-to-rural gradient in relation to household characteristics, their residential location preferences, and travel behaviors which affect land cover and transportation emissions.
  • Hypothesis 3: The relationships between land use and carbon stocks and fluxes is influenced by natural and land-use and legacies.
  • Hypothesis 4: Urban development choices are sensitive to carbon mitigation policies, but their ability to shape the urban structure is highly dependent on the existing built infrastructure.
Box 1 Mechanisms affecting terrestrial carbon stocks and fluxes along an urban gradient

We identify five key mechanisms that affect change in C stocks and fluxes along a gradient of urbanization: land-cover change, emissions, organic inputs, temperature, and N fertilization

Taken together, we hypothesize that these five mechanisms will produce nonlinear variations in C stocks and fluxes across the urban gradient. The amount of C in vegetative biomass (and soils) is expected to generally increase with decreased development intensity, with a small peak in the older suburbs and exurbs where larger lots have had time to accumulate biomass following initial clearing. Fluxes (per unit mass) might be expected to decrease with decreasing temperatures and decreased N and CO2 fertilization but ultimately be highest in the least dense areas because of the large amount of photosynthetically active vegetation in forests.

Box 2 Case Study of Urban Carbon Stocks along a Gradient of Urbanization in the Seattle Metropolitan Area

Initial observations in the Seattle Metropolitan Area provide insights on how C signatures vary across land-cover types on a gradient of urbanization. We used a stratified random sample of 150 plots, with a radius of 15 m, to estimate aboveground live biomass across the Seattle urbanizing region in five land-cover classes and across three transects Figure 3. We sampled five land-cover types including high-density urban, medium urban, low urban, mixed forest, and coniferous forest; thirty sites were sampled per cover type. We used a Landsat TM (2002) land-cover classification to stratify our field samples (Alberti, Weeks, and Coe 2004a; Alberti et al. 2004b).
We quantified aboveground C stocks and assessed site characteristics within the sampled five different land-cover classes (Hutyra et al. 2011a). Aboveground live biomass across the Seattle region was 89 ± 22 Mg Cha1 per year in 2002 (including both urban and forest area), with an additional 11.8 ± 4 Mg Cha1 of coarse woody debris (CWD) biomass. The average biomass stored within forests and urban covers was 140 ± 40 and 18 ± 13.7 Mg Cha1, respectively (Figure 4).
These results are substantially larger than the 25.1 Mg Cha1 (urban forest land only, including both above- and belowground C) reported by Nowak and Crane (2002) for ten U.S. cities and larger than the average of 53.5 Mg Cha1 for all U.S. forests (urban and rural) reported by Birdsey and Heath (1995). For comparison, the Harvard Forest Long Term Ecological Research (LTER; forest age approximately 100 years, 115 Mg Cha1) is one of the most studied forests, from a C cycle perspective (Urbanski et al. 2007), and it contains less aboveground C than the forested land covers within the Seattle urbanizing region.
The remarkable magnitude of observed C stocks in the rapidly urbanizing Seattle region is particularly clear when compared to C stored in the forest cover (even within the urban core) to the biomass stored in Amazonian rainforests. The regional conifer forests stored an average of 182 ± 60 Mg Cha1, which is comparable to the 197 ± 11.6 Mg Cha1 aboveground live C stocks reported for a well-studied, primary Amazonian rainforest (Pyle et al. 2008).
Using a time series analysis of land cover, Hutyra et al. (2011b) also explored the aboveground C stock patterns over two decades (1986 to 2007) in the Seattle Metropolitan Statistical Area (Figure 5). Land-cover change contributed an average annual loss of 1.2 Mg Cha1 in terrestrial C stores. These vegetative C losses corresponded to nearly 15 percent of the regional fossil fuel emissions.


Urban regions are rapidly expanding, and in many respects, urban expansion is critical to advancing viable scenarios for net emissions reduction. How patterns of development mediate their impact could provide important information for to build a robust urban planning and management strategy. Empirical data that accurately take into account the diverse sources and sinks of C in urban regions are critical to gain a mechanistic understanding of the urban C cycle and to establish relationships with patterns of urbanization. Empirical data show remarkable magnitudes of C stocks in the rapidly urbanizing Seattle region and a complex relationship between land use and land cover across the urban to rural gradient (Hutyra et al. 2011). Monitoring C fluxes by installing flux towers across varied metropolitan settings has the potential to advance such strategies. In addition we will need to develop scenarios to explore the most divergent points of uncertainty and important trajectories of land use land cover change. Through a fusion of observations and increased mechanistic understanding, we can use scenarios to explore alternative futures and test alternative policy strategies to tackle the C challenge.