Emission Scenarios and Global Climate Models
Global warming is caused primarily by heat-trapping emissions released into the atmosphere when we burn fossil fuels (to drive our cars or generate electricity in power plants) and cut down forests. To assess how temperature, precipitation, and other aspects of our climate may change with global warming and how these changes might in turn affect specific sectors such as agriculture, forests, or fisheries, we start by identifying any changes that have already happened. We then use global climate models to project what climate changes we might expect in the future and, finally, use these projected changes as input into research on specific consequences within sectors.
Projections of future emissions. To estimate future changes that might be expected in our climate, we first have to make some assumptions about what the future might look like, especially the amount of heat-trapping emissions that will be in the atmosphere. Scientists make estimates of future emissions by creating scenarios of what population growth, energy use, economic development, and technology use could look like in the future.
Scientists and other technical experts involved with the Intergovernmental Panel on Climate Change (IPCC) have developed a set of future emission scenarios known collectively as SRES (Special Report on Emissions Scenarios). We use the SRES A1fi and the B1 scenarios to represent possible higher- and lower-emission choices, respectively, over the coming century (see the figure).
Projected future carbon emissions for the SRES emission scenarios (Nakićenović et al., 2000). The higher-emission scenario (A1fi) corresponds to the highest red dotted line, while the lower-emission (B1) scenario is indicated by the solid green line.
The higher-emission scenario (A1fi, uppermost dotted red line) represents a future with fossil fuel-intensive economic growth and a global population that peaks mid-century and then declines. In this scenario, concentrations of atmospheric carbon dioxide (the main heat-trapping gas) reach 940 parts per million (ppm) by 2100—more than triple pre-industrial levels.
The lower-emission scenario (B1, solid green line) also represents a world with high economic growth and a global population that peaks by mid-century, then declines. However, the lower-emission scenario includes a shift to less fossil fuel-intensive industries and the introduction of clean and resource-efficient technologies. Atmospheric carbon dioxide concentrations reach 550 ppm by 2100, about double pre-industrial levels. Current carbon dioxide concentrations stand at 380 ppm (about 40 percent above pre-industrial levels).
Projections of future changes. These emission scenarios are then used in global climate models to project a range of possible changes. Global climate models incorporate the latest scientific understanding of the physical processes at work in the atmosphere, oceans, and Earth's surface—and how they are all interconnected. A global climate model can produce projections of precipitation, temperature, pressure, cloud cover, humidity, and a host of other climate variables for a day, a month, or a year.
Models were selected for use in the research based on a rigorous set of criteria, including the model's effectiveness in reproducing past and current climate within our region. If a model can replicate known historical conditions in the Northeast, we have higher confidence when projecting the future climate. While there are many global climate models available, the analyses described on this website used a core set of models that performed best.
Because the models are global, they can "wash out" finer distinctions at the regional scale. Grid cells can range in size from 50 to 250 miles per side but, in reality, people don't live globally; they live in a particular place—a state or region. To be meaningful for people's lives, the global model projections must then be adjusted down to a more regional scale. These reports use several well-regarded downscaling techniques to transform global climate model results into projections based on tens of miles rather than hundreds.