Contrasting internally and externally generated Atlantic Multidecadal Variability and the role for AMOC in CMIP6 historical simulations.

Atlantic multidecadal variability (AMV) has long been thought to be an expression of low-frequency variability in the Atlantic Meridional Overturning Circulation (AMOC). However, alternative hypotheses have been forwarded, including that AMV is primarily externally forced. Here, we review the current state of play by assessing historical simulations made for the sixth coupled model intercomparison project (CMIP6). Overall, the importance of external forcing is sensitive to the type of AMV index used, due to the importance of globally coherent externally forced signals in the models. There are also significant contrasts between the processes that drive internally and externally forced AMV, but these processes can be isolated by exploring the multivariate expression of AMV. Specifically, internal variability in CMIP6 models is consistent with an important role of ocean circulation and AMOC and the externally forced AMV is largely a surface-flux forced mechanism with little role for the ocean. Overall, the internal multivariate fingerprint of AMV is similar to the observed, but the externally forced fingerprint appears inconsistent with observations. Therefore, climate models still suggest a key role for ocean dynamics, and specifically AMOC, in observed AMV. Nevertheless, models remain deficient in a number of areas, and a stronger role for externally forced dynamical changes cannot be ruled out. This article is part of a discussion meeting issue 'Atlantic overturning: new observations and challenges'.

Challenges simulating the AMOC in climate models.

The latest assessment report from the Intergovernmental Panel on Climate Change concluded that the Atlantic Meridional Overturning Circulation (AMOC) was very likely to decline over the twenty-first century under all emissions scenarios; however, there was low confidence in the magnitude of the decline. Recent research has highlighted that model biases in the mean climate state can affect the AMOC in its mean state, variability and its response to climate change. Hence, understanding and reducing these model biases is critical for reducing uncertainty in the future changes of the AMOC and in its impacts on the wider climate. We discuss how model biases, in particular salinity biases, influence the AMOC and deep convection. We then focus on biases in the UK HadGEM3-GC3-1 climate model and how these biases change with resolution. We also discuss ongoing model development activities that affect these biases, and highlight priorities for improved representation of processes, such as the position of the North Atlantic Current, transports in narrow boundary current, resolution (or improved parameterization) of eddies and spurious numerical mixing in overflows. This article is part of a discussion meeting issue 'Atlantic overturning: new observations and challenges'.

On the risk of abrupt changes in the North Atlantic subpolar gyre in CMIP6 models.

CMIP5 models have been shown to exhibit rapid cooling events in their projections of the North Atlantic subpolar gyre. Here, we analyze the CMIP6 archive, searching for such rapid cooling events in the new generation of models. Four models out of 35 exhibit such instabilities. The climatic impacts of these events are large on decadal timescales, with a substantial effect on surface temperature over Europe, precipitation pattern in the tropics-most notably the Sahel and Amazon regions-and a possible impact on the mean atmospheric circulation. The mechanisms leading to these events are related to the collapse of deep convection in the subpolar gyre, modifying profoundly the oceanic circulation. Analysis of stratification in the subpolar gyre as compared with observations highlights that the biases of the models explain relatively well the spread in their projections of surface temperature trends: models showing the smallest stratification biases over the recent period also show the weakest warming trends. The models exhibiting abrupt cooling rank among the 11 best models for this stratification indicator, leading to a risk of encountering an abrupt cooling event of up to 36.4%, slightly lower than the 45.5% estimated in CMIP5 models.

Natural drivers of multidecadal Arctic sea ice variability over the last millennium.

The climate varies due to human activity, natural climate cycles, and natural events external to the climate system. Understanding the different roles played by these drivers of variability is fundamental to predicting near-term climate change and changing extremes, and to attributing observed change to anthropogenic or natural factors. Natural drivers such as large explosive volcanic eruptions or multidecadal cycles in ocean circulation occur infrequently and are therefore poorly represented within the observational record. Here we turn to the first high-latitude annually-resolved and absolutely dated marine record spanning the last millennium, and the Paleoclimate Modelling Intercomparison Project (PMIP) Phase 3 Last Millennium climate model ensemble spanning the same time period, to examine the influence of natural climate drivers on Arctic sea ice. We show that bivalve oxygen isotope data are recording multidecadal Arctic sea ice variability and through the climate model ensemble demonstrate that external natural drivers explain up to third of this variability. Natural external forcing causes changes in sea-ice mediated export of freshwater into areas of active deep convection, affecting the strength of the Atlantic Meridional Overturning Circulation (AMOC) and thereby northward heat transport to the Arctic. This in turn leads to sustained anomalies in sea ice extent. The models capture these positive feedbacks, giving us improved confidence in their ability to simulate future sea ice in in a rapidly evolving Arctic.

Limits on determining the skill of North Atlantic Ocean decadal predictions.

The northern North Atlantic is important globally both through its impact on the Atlantic Meridional Overturning Circulation (AMOC) and through widespread atmospheric teleconnections. The region has been shown to be potentially predictable a decade ahead with the skill of decadal predictions assessed against reanalyses of the ocean state. Here, we show that the prediction skill in this region is strongly dependent on the choice of reanalysis used for validation, and describe the causes. Multiannual skill in key metrics such as Labrador Sea density and the AMOC depends on more than simply the choice of the prediction model. Instead, this skill is related to the similarity between the nature of interannual density variability in the underlying climate model and the chosen reanalysis. The climate models used in these decadal predictions are also used in climate projections, which raises questions about the sensitivity of these projections to the models' innate North Atlantic density variability.