We address this need by investigating the GHG emissions and land use impacts of dramatically different urbanization storylines1 for an agricultural county within one of California’s rapidly growing metropolitan regions. In contrast to traditional urban-growth modeling, which projects scenarios into the future based on current and past policy, we take a “back casting” approach that seeks to consider the implications of radically different alternative strategies at a date far in the future. Accordingly, we propose strongly different story lines for 2050, develop modeling parameters based on these alternative futures, model the spread of new urban development across the landscape between now and then, and then estimate annual emissions from transportation and residential energy use in 2050 as well as the effects of urban growth on agricultural land. In terms of urbanization, our story lines range from business as usual in the county, with 65% of new households in traditional suburban or exurban densities, to a very-compact development scenario with only 10% of new households in these categories. Also factored into the scenarios are differential levels of urban rural connectivity such as local food marketing and consumption, which help build interest and support in the climate-related co-benefits of agriculture. The results are necessarily broad-brush, given that population, economic conditions, and political attitudes cannot be estimated with any degree of precision over such a long period. Still, such an approach can be useful to illustrate dramatically different policy approaches,cut flower transport bucket and indeed can be seen as necessary in order to give policymakers and the public a sense of the level of change required to meet climate change planning goals . As a foundation for our work, we use the Special Report on Emissions Scenarios storylines that the Intergovernmental Panel on Climate Change established in 2000.

According to the IPCC working group, “Scenarios are alternative images of how the future might unfold and are an appropriate tool with which to analyze how driving forces may influence future emission outcomes and to assess the associated uncertainties” . The IPCC scenarios are based on very broad storylines for alternative global futures, specifying different trajectories for population, globalization, economic growth, and environmental protection. The working group intended them to assist in the modeling of future GHG emissions, and also to assist with understanding of global warming impacts, climate adaptation , and mitigation . We chose the A2 and B1 scenarios for higher and lower GHG emissions, respectively, which can be conceptually down scaled to explore how future local land use patterns will respond to climate change . Scenario A2 has higher economic and population growth and less emphasis on sustainability priorities than Scenario B1. Because IPCC storylines do not include specific action to mitigate GHG emissions, we have added a third alternative, called AB32-Plus, which assumes continued development of State of California climate change policy as set out by a 2006 law, Assembly Bill 32 , as well as other state legislation and policy. In particular, Senate Bill 375 of 2008 requires each metropolitan area to develop a Sustainable Communities Strategy aimed at coordinating land use and transportation planning so as to reduce GHG emissions from transportation. Although coordinated transportation–land use planning in California certainly predates these pieces of climate change legislation , the state’s metropolitan areas began developing the new Sustainable Communities Strategies in the early 2010s , potentially establishing a stronger trajectory of urban growth planning. We sought to design urbanization assumptions within the AB32-Plus storyline so as to meet the state’s GHG mitigation goals as well as to achieve other benefits such as farmland preservation, greater provision of ecosystem services at the rural–urban interface, biodiversity conservation, improved rural livelihood options, and business opportunities that build social capital .

Our overall process then, was to review relevant literature, assemble storylines and scenario assumptions, model urbanization for the county with geographic information system–based software, calculate likely transportation and building emissions from new residential development for each scenario,and assess land use change implications. The analysis concludes with a number of strategies, some already in progress, which could inform a growth-management framework to limit urban development and enhance preservation of agricultural lands.However, Solecki and Oliveri used this strategy to examine conversion of agricultural to urban land in the New York City area, employing the SLEUTH urban-growth model to investigate A2 and B2 trend scenarios for 2020 and 2050, with 1960–1990 growth as a base. Modeling parameters primarily concerned the ways urban grid cells propagated in relation to existing development, urban edges, and transportation infrastructure. Solecki and Oliveri’s B2 scenario was substantially weaker in managing urban growth than the alternatives we envisioned developing. Although these authors found less urban sprawl with their environmentally oriented alternative, the percentage of land urbanized still more than tripled from 1990 to 2050. Rounsevell et al. Down scaled four SRES storylines for Europe and modeled land use for 2020, 2050, and 2080 time frames, though at a much larger spatial scale than ours . The main drivers for their model were global resource, market, and policy assumptions rather than local land use policy. Not surprisingly, their relatively green B1 and B2 scenarios performed best at preserving agricultural lands. Barredo and Gómez tested a cellular automata–based model through analysis of urban growth on 10,000 square kilometers around Madrid under three SRES scenarios for the 2000–2040 period. Model parameters focused on land accessibility, suitability, zoning status, and neighborhood effects. Their method produced distinctly different spatial clustering and distribution of development for their different storylines. Van Eck and Koomen applied two scenarios based on SRES storylines to model urban concentration and land use diversity in the Netherlands, finding that the latter produced significantly more urban sprawl and less concentration of development. None of these researchers, though, sought to further link their models to GHG emissions from the predicted development patterns.

More general analysis of urban growth has long supported the supposition that low-density suburban sprawl increases motor vehicle use, leading to higher GHG emissions compared with non-urban uses on the same land or with similar new populations living in denser urban environments with greater land use diversity . In a 2009 review of the literature, the National Research Council concluded that doubling residential density across a metropolitan area, combined with improved land use mix and transit, might lower household vehicle miles traveled by 5% to 12%, and perhaps by as much as 25% . The relationship is complex, however . In an analysis of 80 growth-scenario planning exercises in 50 US regions, Bartholomew attributed the relatively modest decreases in VMT usually shown within compact-development scenarios to the traditional insensitivity of travel-demand models to land use patterns,procona flower transport containers as well as the omission of other variables such as land use diversity and pricing. Sheer population and job densities may not be as important as residents’ accessibility to destinations and street-network design . Other factors such as the availability of public transit, bicycle and pedestrian infrastructure, and economic incentives probably play important roles as well.Research on relationships between urban form and GHG emissions is still in the early stages, and is based primarily on modeled rather than observed data. Andrews developed an exploratory land use–GHG emissions analysis framework that considers emissions from buildings, transportation, waste management, landscape management, urban heat islands, and electricity transmission and distribution. Applying this framework to typical types of development found in New Jersey towns, he found per capita CO2 emissions varying by a factor of two, with transportation emissions much lower in dense urban locations than in suburban ones, building emissions somewhat lower, and single family detached homes producing 33% more GHG from energy use than units in multifamily structures. Carbon sequestration within forests substantially lowered per capita human emissions in exurban locations compared with suburban or urban settings around the periphery of these towns in this East Coast location. This is less likely to be important in arid or primarily agricultural areas of the country, where the amount of woody vegetation is much lower. Waste management, urban heat-island effects, and electric transportation and distribution losses all proved relatively small factors in Andrews’s analysis. Ewing and Rong investigated the relation between suburban sprawl and residential-building energy consumption, finding 54% higher energy consumption for space heating for single-family detached units when compared with similar households in multifamily structures. However, they also found that urban areas have somewhat increased energy consumption for cooling, due to urban heat-island effects.

In a study of relatively low-density vs. high-density neighborhoods in the Toronto area, Norman, MacLean, and Kennedy found GHG emissions from the former approximately 81% higher for building operations and 365% higher for transportation activities. In a study of 11 metropolitan regions in the Midwestern US, Stone et al. estimated that an aggressive smart-growth scenario over 50 years could reduce the growth in transportation emissions from business-as-usual development by 34%, and that over business as usual, and that this land use strategy, combined with use of hybrid-electric vehicles, could reduce the growth in emissions by 97%. The relation between transportation emissions and building-related emissions will vary according to climate and geographical region. Randolph believes that in general sprawl has far greater impacts on transportation emissions than on building emissions. However, Andrews points out that, in some locations at least, building emissions are greater in quantity. Although the idea of modeling urban growth with very-low-GHG scenarios has been rare in academia, public agencies are beginning to move toward such back casting approaches in an effort to meet emissions-reduction targets and related legislation. As mentioned previously, California’s 2008 SB 375 legislation begins the process of encouraging such scenarios throughout California. The Sacramento Area Council of Governments , within the preparatory work for its 2012 Sustainable Communities Strategy , developed two significantly different future scenarios for the region based on different assumed energy efficiencies . Apparently, neither spatially explicit modeling of urbanization nor the official Sustainable Community Strategy were included, but this nevertheless represents a relatively strong environmentally oriented urban-growth vision given the current politics around land use. In fact, the Sustainable Community Strategy is not highly conducive to a carbon-neutral future, given that more than 25% of the region’s new housing in 2035 would continue to be built in the form of large-lot single-family homes outside of existing urbanized areas , adding to the region’s large existing stock of such homes. Modeling of the agency’s land use scenario, together with revised transpor-tation priorities, reduces transportation-related GHG emissions by 20% by 2020 compared to 2008, but emissions-reduction progress stalls thereafter, producing only an additional 3% improvement by 2035, far short of the trajectory needed for the state’s 2050 emission-reduction goal . If land use is to contribute toward meeting long-term GHG-reduction targets, dramatically different scenarios appear necessary.Yolo County is generally representative of agricultural counties in California’s Central Valley in that it contains a mix of irrigated perennial and row crops on alluvial plains, upland grazed rangelands, and small towns and cities. These agricultural landscapes also contain riparian corridors and other types of wetlands that are important for natural resource and biodiversity conservation. The Central Valley is one of the most productive agricultural regions in the world, yet is facing some of the most rapid population growth in the state. Urbanization in Yolo County is somewhat slower than in many other areas of the valley, having fallen to approximately 1% annually during the economic slowdown starting in the late 2000s. Total population was 200,709 in 2009; predictions for 2050 range from 320,000 to 394,000 . Given the county’s geography, urban expansion will almost certainly occur at the expense of farmland and open space if growth is not restricted to infill development within existing urban boundaries. Yolo County includes 653,452 acres of land . The incorporated cities of Davis, West Sacramento, Woodland, and Winters account for about 4.6% of the land area . In 1998, Yolo County alone contained about 40% of the prime farmland in the Sacramento region and yielded the highest farm market values out of all the counties . Thus, the jurisdiction is an important reservoir of productive farmland within an urban region.