Accurate climate information, such as future projections, is needed on relevant spatial and temporal scales to inform user’s decision making needs. However, using projections to assess climate change risks remains challenging, in part due to the large associated uncertainties, such as those for extreme events.
CPO’s ESSM Division’s programs including MAPP, CVP, and COM contribute to U.S. CLIVAR and support topical expert working groups, such as the Large Ensembles Working Group, that are addressing these challenges. This week, the Large Ensembles Working Group published a perspective in Nature Climate Change, demonstrating the value of a collection of large ensembles generated with seven Earth system models, the Multi-Model Large Ensemble Archive (MMLEA), for providing new insights into addressing uncertainties. The paper also highlights the Archive’s potential use to stimulate new research directions and Earth system applications. The recent availability of the Community Earth System Model version 1- Large Ensemble (CESM1-LE) as a public dataset on Amazon Web Services cloud is evidence of broad interest and demand for these types of data, representing “truly a sea change for climate and related sciences”.
A key discerning element of the Archive is that it can be used to separate uncertainties due to natural internal variability in the coupled ocean-atmosphere-land-biosphere-cryosphere system versus those due to structural model formulation differences. This is important because, unlike internal natural variability, structural model difference-based uncertainties will improve as models improve.
The MMLEA will yield “new insights on separating sources of uncertainties”. For example, it allows for the separation of sources of uncertainty at smaller spatial and temporal scales, and for notoriously variable quantities such as precipitation and temperature. MMLEA is particularly relevant for “decision-making and risk assessment in a variable climate” since internal variability has a large impact on rare, extreme events. Additionally, MMLEA can be used as a “methodological testbed for observations”. Opportunities exist to take advantage of historical and paleoclimate records to effectively evaluate and benchmark internal variability generated by model large ensembles.
Authors note the MMLEA has a number of “emerging Earth system applications” for societal and science needs. A subset of stakeholders are already well-positioned to harness the power of the existing MMLEA to further public health, agriculture, and fishery applications. However, there is likely much untapped potential in broader societal applications for ecosystem management, food security, and public health. Future development of MMLEs for systems where internal natural variability is high under anthropogenic forcing, such as atmospheric chemistry or ocean biogeochemistry, could further new research directions with clear societally-relevant outcomes – such as those that would inform air quality and public health planning and/or optimization of observing system design and duration. The authors call for continued efforts to make these data accessible to and useable by a broad community.