Atmospheric Sciences Meeting Report

Developing Climate Model Comparisons

Experimental Design for CMIP6: Aerosols, Land Use, and Future Scenarios;
Aspen, Colorado, 4–8 August 2014

By Brian C. O’Neill, Jean-François Lamarque, and David Lawrence

Since 1995, the worldwide climate modeling community has designed and participated in a series of intercomparison projects for the purposes of understanding and improving model performance, investigating scientific questions about the climate system, and projecting future climate conditions. These projects have been defined under the umbrella of the Coupled Model Intercomparison Project, and phase 6 of that project (CMIP6) is just getting under way. As in previous phases, many CMIP6 modeling activities interact and overlap with each other. For example, credible projections of future climate conditions require understanding and validating a variety of Earth system model responses, including those to changes in concentrations of greenhouse gases, aerosols and other air pollutants, and land use change.

To account for these overlaps, scientists involved in three potential CMIP6 modeling activities met together in August at the Aspen Global Change Institute in Aspen, Colo., to further develop proposed model comparisons on the topics of future scenarios (­ScenarioMIP), land use (LUMIP), and atmospheric chemistry and aerosols (­AerChemMIP). A primary goal of the workshop was to identify and discuss key crosscutting scientific questions and to develop complementary experimental designs to address them.

The goal of ScenarioMIP is to produce climate model simulations of plausible alternative future forcing pathways that will underpin integrated analysis of climate change impacts as well as mitigation and adaptation responses. Meeting participants identified a small set of scenarios that would serve this purpose and would update and expand on scenarios that were run by climate models in previous model comparisons. In addition, to better understand how sensitive the projected future climate might be to land use or regional air pollutant emissions, specific scenarios were identified that will serve as the basis for sensitivity experiments in the other comparison projects.

LUMIP activities are designed to assess the effects of land use and land use change on climate and biogeochemical cycling, with a new emphasis on the effects of different land management practices (e.g., wood harvest, irrigation, and fertilization). LUMIP data standardization efforts will generate updated and expanded historic land use (including management) data sets, as well as projected future trends. Land use modelers also designed a factorial set of land-only experiments, with each experiment either adding or removing a modeled land use feature, such as prognostic crop modeling, human fire management, or wood harvest.

The AerChemMIP activities are designed to diagnose forcings and feedbacks of near-term climate forcers (namely, tropospheric aerosols, ozone, and methane) and ozone-depleting substances in the upcoming CMIP6 simulations. This diagnosis requires documenting and understanding past and future changes in the chemical composition of the atmosphere (including regional air quality) and estimating the climate impacts of these changes. Meeting participants identified an initial set of experiments and diagnostics to achieve these objectives, including perturbation simulations from historical or projection experiments.

The workshop contributed to the development of proposals for these activities to become part of CMIP6. Members working on each of these activities will be refining proposals over the coming months, with the aim of finalizing experimental designs by 31 March 2015.

—Brian C. O’Neill, Jean-Francois Lamarque, and David Lawrence, National Center for Atmospheric Research, Boulder, Colo.; email: [email protected]


Citation: O’Neill, B. C., J.-F. Lamarque and D. Lawrence (2014), Developing climate model comparisons, Eos Trans. AGU, 95(49), 462, doi:10.1002/2014EO490008.

© 2014. American Geophysical Union. All rights reserved.