Drivers of Low-Frequency Sahel Precipitation Variability: Comparing CMIP5 and CMIP6 with Observations

Rebecca Jean Herman,^a Michela Biasutti,^b Yochanan Kushnir^b

^a Department of Earth and Environmental Sciences of Columbia University, New York, NY

^b Lamont-Doherty Earth Observatory of Columbia University, Palisades, NY

Corresponding author : Rebecca Herman, rebecca.herman@columbia.edu

ABSTRACT

We examine and contrast the simulation of Sahel rainfall in phases 5 and 6 of the Coupled Model Intercomparison Project (CMIP5 and CMIP6). On average, both ensembles grossly underestimate the magnitude of low-frequency variability in Sahel rainfall. But while CMIP5 partially matches the timing and pattern of observed multi-decadal rainfall swings in its historical simulations, CMIP6 does not. To classify model deficiency, we use the previously-established link between changes in Sahelian precipitation and the North Atlantic Relative Index (NARI) for sea surface temperature (SST) to partition all influences on Sahelian precipitation into five components: (1) teleconnections to SST variations; the effects of (2) atmospheric and (3) SST variability internal to the climate system; (4) the SST response to external radiative forcing; and (5) the “fast” response to forcing, which is not mediated by SST. CMIP6 atmosphere-only simulations indicate that the fast response to forcing plays only a small role relative to the predominant effect of observed SST variability on low-frequency Sahel precipitation variability, and that the strength of the NARI teleconnection is consistent with observations. Applying the lessons of atmosphere-only models to coupled settings, we imply that the failure of coupled models in simulating 20^th century Sahel rainfall derives from their failure to simulate the observed combination of forced and internal variability in SST. Yet differences between CMIP5 and CMIP6 Sahel precipitation do not mainly derive from differences in NARI, but from either their fast response to forcing or the role of other SST patterns.

1. Introduction

The semi-arid region bordering the North African Savanna and the Sahara Desert, known as the Sahel, received much scientific attention since it experienced unparalleled dramatic rainfall variability in the second half of the 20^th century. The importance of teleconnections between Sahel precipitation and global sea surface temperature (SST) was demonstrated in the early stages of Sahel climate variability research (Folland et al. 1986; Giannini et al. 2003; Knight et al. 2006; Palmer 1986; Zhang and Delworth 2006), and has been further reinforced in more recent studies (Okonkwo et al. 2015; Parhi et al. 2016; Park et al. 2016; Pomposi et al. 2015; Pomposi et al. 2016; Rodríguez-Fonseca et al. 2015 and references therein). But while the dominant role of SST in driving the pacing (though not necessarily the full magnitude) of 20th century Sahel rainfall variability is unquestioned (Biasutti 2019), there is still debate on whether the evolution of SST and the related Sahel precipitation variability were externally forced (Ackerley et al. 2011; Biasutti 2013; Biasutti and Giannini 2006; Biasutti et al. 2008; Bonfils et al. 2020; Dong and Sutton 2015; Giannini and Kaplan 2019; Haarsma et al. 2005; Haywood et al. 2013; Held et al. 2005; Hirasawa et al. 2020; Hua et al. 2019; Iles and Hegerl 2014; Kawase et al. 2010; Marvel et al. 2020; Polson et al. 2014; Undorf et al. 2018; Westervelt et al. 2017) or the manifestation of variability internal to the climate system (IV, Sutton and Hodson 2005; Ting et al. 2009; Zhang and Delworth 2006).

Recently, Herman et al. (2020, hereafter H20) investigated multi-model means (MMM) of historical simulations from the Coupled Model Intercomparison Project phase 5 (CMIP5, Taylor et al. 2012), and found that anthropogenic aerosols (AA) and volcanic aerosols (VA), but not greenhouse gases (GHG), were responsible for forcing simulated Sahelian precipitation that correlates well with observations, with AA alone responsible for the low-frequency component of simulated variability. This conclusion appeared consistent with previous claims that AA emissions, which increased until the 1970s and then decreased in response to clean air initiatives (Klimont et al. 2013; Smith et al. 2011), caused multi-decadal variability in Sahel precipitation via changes in Northern Hemisphere surface temperature (Ackerley et al. 2011; Haywood et al. 2013; Hwang et al. 2013; Undorf et al. 2018), or specifically via multidecadal variability in North Atlantic SST (the Atlantic Multidecadal Variability, AMV; Booth et al. 2012; Hua et al. 2019). However, H20 also found that the simulated rainfall response to forcing has little low-frequency power relative to observations, and that simulated IV is unable to account for this difference.

H20 and most other attribution studies do not examine in depth the pathways through which AA (and for that matter, IV and other external forcing agents) affect Sahel precipitation. Thus, H20 did not determine whether the discrepancy between CMIP5 simulations and observations represents an underestimate of aerosol indirect effects and climate feedbacks that amplify the simulated precipitation response to AA, or a fundamental inability of the models to simulate aspects of the observed climate response to forcing or observed modes of IV. Identifying the deficiencies in model representation of the pathways by which external forcing and IV influence the West African Monsoon and Sahel rainfall is essential for attribution of 20^th century changes and also for prediction of this region’s climate future, as model simulations don’t even agree on the sign of future precipitation changes in the Sahel (Biasutti 2013).

Here, we use the well-established link between SST and Sahel precipitation to decompose the effects of individual external forcing agents (F) and internal variability (IV) on Sahel precipitation (P) into five path components, presented in Figure 1: (1) teleconnections that communicate variations in SST to variations in P (indicated by the arrow\(\overrightarrow{t}\)); (2) the “fast” atmospheric and land-mediated effect of external forcing (F) on P (\(\overrightarrow{f}\)); (3) the direct effect of atmospheric IV on P (\(\overrightarrow{a}\)); (4) the effect of F on SST (\(\overrightarrow{s}\)); and (5) the impact of IV in the coupled climate system on SST (\(\overrightarrow{o}\)). The path\(F\rightarrow SST\rightarrow P\) is the “slow,” SST-mediated effect of F on P.