Geophysical Research Letters

Inability of stratospheric sulfate aerosolinjections to preserve the WestAntarctic Ice SheetK. E. McCusker1,2, D. S. Battisti1, and C. M. Bitz1

1Department of Atmospheric Sciences, University of Washington, Seattle, Washington, USA, 2Now at School of Earth andOcean Sciences, University of Victoria, Victoria, British Columbia, Canada

Abstract Injection of sulfate aerosols into the stratosphere has the potential to reduce the climateimpacts of global warming, including sea level rise (SLR). However, changes in atmospheric and oceaniccirculation that can significantly influence the rate of basal melting of Antarctic marine ice shelves and theassociated SLR have not previously been considered. Here we use a fully coupled global climate model toinvestigate whether rapidly increasing stratospheric sulfate aerosol concentrations after a period of globalwarming could preserve Antarctic ice sheets by cooling subsurface ocean temperatures. We contrast thisclimate engineering method with an alternative strategy in which all greenhouse gases (GHG) are returnedto preindustrial levels. We find that the rapid addition of a stratospheric aerosol layer does not effectivelycounteract surface and upper level atmospheric circulation changes caused by increasing GHGs, resultingin continued upwelling of warm water in proximity of ice shelves, especially in the vicinity of the alreadyunstable Pine Island Glacier in West Antarctica. By contrast, removal of GHGs restores the circulation,yielding relatively cooler subsurface ocean temperatures to better preserve West Antarctica.

1. Introduction

The retreat of the Pine Island and Thwaites Glaciers has accelerated, with a rapid collapse of the West AntarcticIce Sheet (WAIS) possible within as little as a few centuries [Joughin et al., 2014; Rignot et al., 2014; Favier et al.,2014]. Estimates of the current WAIS mass contribution to sea level are relatively small (about 4 mm for theperiod 1992–2011) [Shepherd et al., 2012]; however, a full collapse could yield sea level rise (SLR) as greatas 4.3 m [Church et al., 2013], exceeding projected steric SLR for the next five centuries [Church et al., 2013].Given that the retreat of Pine Island and Thwaites Glaciers may be slowed or reversed if basal melt rates aredecreased [Favier et al., 2014], an important question is whether solar radiation management (SRM) may beable to delay or even prevent a WAIS collapse, as has been suggested [Blackstock et al., 2009].

A large amount of mass loss from ice sheets occurs due to basal melt of the ice shelves [Joughin and Alley,2011], the floating extensions of land ice sheets. In West Antarctica, the faces of these ice shelves extendfrom the surface to several hundred meters depth, near the level of the Circumpolar Deep Water (CDW), asubsurface water mass that is slightly warmer (−1∘C to 1∘C) [Yin et al., 2011] than waters above and belowit. An increase in the temperature of CDW or in the rate that CDW is brought onto the continental shelf dueto regional wind variability can greatly increase basal melting of ice shelves [Thoma et al., 2008; Joughin andAlley, 2011], causing ice shelf thinning. While the thinning of floating ice shelves does not in itself affectthe sea level, the subsequent reduction in buttressing stress leads to an increase in ice stream velocity andice mass loss, indirectly leading to global average SLR [Oppenheimer, 1998; Pritchard et al., 2012]. Thus, iceshelf disintegration and the subsequent SLR can happen even in the absence of large atmospheric warming[Oppenheimer, 1998].

SRM has been shown to reduce the rate of global mean sea level rise. By examining observed relationshipsbetween sea level and volcanic eruptions, which temporarily lower sea level by briefly reducing ocean heatcontent [Church et al., 2005; Gleckler et al., 2006], Moore et al. [2010] argued that sulfate aerosol injections asSRM would reduce the rate of global mean sea level rise. Irvine et al. [2012] show that steric height changesin an intermediate complexity climate model are reduced by insolation reductions as SRM, and additionallyinclude the contribution to sea level from the mass input of glaciers, icecaps, and Greenland and Antarcticice sheets with a scaling to the global mean surface temperature anomaly. However, changes in oceanic

RESEARCH LETTER10.1002/2015GL064314

Key Points:• Sulfate aerosol injection allows warm

Southern Ocean water to reachice shelves

• Risk of Antarctic ice sheet instabilitypossible without surface warming

• Quantitative assessment of sea levelrise under aerosol injections needed

Supporting Information:• Figures S1–S3

Correspondence to:K. E. McCusker,kemccusk@uvic.ca

Citation:McCusker, K. E., D. S. Battisti, andC. M. Bitz (2015), Inability ofstratospheric sulfate aerosolinjections to preserve the WestAntarctic Ice Sheet, Geophys. Res. Lett.,42, doi:10.1002/2015GL064314.

Received 21 APR 2015

Accepted 2 JUN 2015

Accepted article online 4 JUN 2015

©2015. American Geophysical Union.All Rights Reserved.

MCCUSKER ET AL. STRATOSPHERIC AEROSOLS AND WAIS 1

Geophysical Research Letters 10.1002/2015GL064314

circulation that can have significant impacts on basal melting of ice shelves [Steig et al., 2013; Joughin andAlley, 2011; Thoma et al., 2008] have not previously been considered.

The subsurface Southern Ocean temperature structure is strongly influenced by the surface wind stress[Fyfe et al., 2007; Spence et al., 2014], which in turn is set by the circumpolar westerly jet. McCusker et al. [2012]showed in a set of idealized simulations that when the global average temperature is stabilized using a strato-spheric sulfate layer that counteracts growing CO2 concentrations, an anomalous poleward intensified zonalwind stress on the Southern Ocean remains that causes upwelling of warmer subsurface waters to the level ofice sheet outlets. Warmed subsurface ocean waters that destabilize marine ice sheets may introduce thresh-old behavior in ice sheet mass loss [Notz, 2009] such that ice sheet loss, and hence SLR, becomes nonlinearwith increased warming [Joughin et al., 2014].

To promptly avoid SLR in the face of rising greenhouse gases (GHGs), a large, rapid input of stratosphericaerosols would need to be deployed to reverse heat uptake by the ocean. Any delays in implementation wouldallow additional ocean heat storage that may linger at intermediate depths for centuries [Gillett et al., 2011].SRM is most likely to be deployed only after a period of global warming that is deemed to be unacceptablylarge. Could a subsequent implementation of sulfate SRM preserve Antarctic ice sheets? Here we investigatewith the Community Climate System Model, version 4 (CCSM4) the impact of a rapidly increased stratosphericsulfate aerosol burden on atmospheric and oceanic circulation in the proximity of Antarctica. We contrastthese results with an alternative management scenario in which GHGs are abruptly returned to preindustriallevels—the ultimate, if not impractical, climate engineering.

2. Methods

In order to capture the relevant range of climate system time scales and dynamical feedbacks, we use theCCSM4 [Gent et al., 2011], a general circulation model (GCM) with a finite volume 0.9∘ ×1.25∘ resolution atmo-sphere coupled to sea ice, land, and 1∘ full-depth ocean models. The ocean model is a constant volume modelwith reference salinity used to compute salinity changes from freshwater input. Glaciers and ice sheets areallowed up to 1 m of snow water equivalent to accumulate, and the rest is routed directly to the ocean. Giventhat there is very little melt on Antarctica presently, precipitation minus evaporation is approximately equalto runoff. In a warming scenario, however, the snowpack is allowed to melt if it becomes warm enough.

Steric sea level change is computed as follows: 𝜂n = H((𝜌0∕𝜌n) − 1), where 𝜂n is the change in surface ele-

vation at time step n, H is the global mean ocean depth, 𝜌0 is the ‘baseline’ global mean ocean density froma twentieth century simulation (20thC) obtained from the National Center for Atmospheric Research (NCAR)averaged over 1970–1999, and 𝜌n is global mean density at time step n. Ocean density, 𝜌, is primarily modi-fied by changes in ocean temperature but is also affected by changes in salinity. This calculation assumes novolume change due to freshwater input (e.g., precipitation, evaporation, river runoff, melting, and freezing ofsea ice).

Two simulations from NCAR are used as controls: the Coupled Model Intercomparison Project 5“business-as-usual” scenario, called the Representative Concentration Pathway 8.5 simulation (RCP8.5), reach-ing a radiative forcing from greenhouse gas emissions of 8.5 W/m2 above preindustrial levels by 2100, andthe twentieth century simulation (20thC) forced with historical greenhouse gas and aerosol emissions plusvolcanic eruptions. The climate engineering simulations are branched from RCP8.5 at year 2035 when theglobal mean surface air temperature (SAT) is approximately 1∘C greater than that of the end of 20thC average(computed for years 1970–1999). In year 2035, a prescribed stratospheric sulfate aerosol burden is com-menced that is ramped up from zero at a rapid rate of increase. The scenario has 3 years of rapid ramping oftotal annual average sulfate (SO4) burden at 8 teragrams of SO4 per year (Tg/yr), followed by ramping of theburden at 0.67 Tg/yr for the remainder of the simulation (calculated to provide a roughly equal and oppositeradiative forcing to RCP8.5).

The prescribed SO4 burden, as in McCusker et al. [2012] and McCusker et al. [2014], is generated from outputfrom simulations described in Rasch et al. [2008] wherein volcanically sized (0.05, 2.03, and 0.17 mm for drymode radius, standard deviation, and effective radius, respectively) SO2 was injected into the stratosphereover the tropics and allowed to circulate and oxidize to SO4. In the present study, we prescribe an SO4 burdenrather than an SO2 injection rate, so the particle size distribution is fixed. Given evidence that increasing

MCCUSKER ET AL. STRATOSPHERIC AEROSOLS AND WAIS 2

Geophysical Research Letters 10.1002/2015GL064314

Figure 1. Time series of SAT and SLR. Global- mean annual-mean (a) surface air temperature (SAT; ∘C), and (b) steric sealevel rise (SLR; cm) due to changes in ocean density, shown as anomalies from 20thC (1970–1999 average). The Sulfcurve is an ensemble average. Light blue shading in Figure 1a indicates the spread of the Sulf ensemble of simulations.

injection rates may cause increased particle size and faster sedimentation rates [e.g., English et al., 2012], theradiative forcing achieved for a given burden may be overestimated in our simulations.

We conducted an ensemble of four sulfate engineering simulations, initialized from four RCP8.5 ensemblemembers. Throughout the paper, we refer to the ensemble average as “Sulf.” We contrast the sulfate engineer-ing scenario with an idealized representation of complete mitigation and carbon sequestration by conductingone simulation, called GHGrem, in which anthropogenic greenhouse gases (CO2, CH4, N2O, and CFCs 11 and12) are abruptly set to year 1850 levels in the year 2035. Unless otherwise noted, anomalies are presented asthe departures of the average response (Sulf, GHGrem, and RCP8.5) in years 2045–2054 from the 1970–1999average from 20thC.

3. Sea Level Rise and Atmospheric Circulation

Density changes in our scenarios constitute the “steric” contribution to sea level rise (section 2). As CCSM4 (andmost models of its class) lacks dynamical ice sheet behavior, and ice shelves and ice shelf cavities are absent,the contribution of mass input from the WAIS and other Antarctic ice sheets cannot be explicitly calculated.We, instead, focus on the large-scale oceanic temperature anomalies that determine melt rates near theice shelves.

MCCUSKER ET AL. STRATOSPHERIC AEROSOLS AND WAIS 3

Geophysical Research Letters 10.1002/2015GL064314

Figure 2. Zonal average anomalies with height. Annual average, zonalaverage (a) temperature (∘C), and (b) zonal wind (m/s) averaged from2045–2054 in the Sulf ensemble minus the 1970–1999 average from20thC. Hatching indicates statistical significance at the 90% level usingthe Student’s t test for difference of two means.

The time evolution of global mean sur-face air temperature (SAT) for RCP8.5and climate engineering scenarios isshown in Figure 1a. Sulf represents ascenario in which drastic climate engi-neering is required to stop sea levelsfrom further rising, and hence entails aquick and large decrease in net radiativeforcing via a rapid increase in strato-spheric aerosol burden. This decrease inradiative forcing avoids a large amountof warming compared to business asusual (Figure 1a; RCP8.5). The global andannual mean SAT linear trend for the firstdecade following the start of SRM (years2035–2044) is −1.2∘C/decade for theensemble mean, and the land-only trendis even greater at −1.5∘C/decade. Onedecade later (2045–2054), Sulf returnsglobal mean SAT to roughly the end of20thC (1970–1999), with a SAT anomalyof −0.03∘C in Sulf compared to 1.83∘Cfor RCP8.5. A second climate engineeringscenario consisting of an instantaneousremoval of GHGs to preindustrial con-ditions (GHGrem) gives a change in SATthat is broadly equivalent to that of Sulffor the same time period (−0.36∘C).

Global average sea level due to changesin ocean density, in isolation of possiblemass change from dynamical ice loss orexceptional melt (i.e., melt beyond 1 mof snow water equivalent; see section 2),reverses course and starts decliningupon initiation of climate engineering(Figure 1b). Nearly 30 cm of steric sea levelrise is avoided by century’s end when asulfate layer is implemented, consistentwith previous findings [Irvine et al., 2012].The relatively long timescale of oceanheat uptake means that by 2095, the sealevel in Sulf is still not fully equilibrated to

the negative radiative forcing imposed by the increasing aerosols and thus, further sea level decline is in thepipeline. Nevertheless, a high rate of sulfate injection could promptly reduce or avoid a substantial amountof steric sea level rise, at the expense of potentially hazardous, large SAT trends. Removal of GHGs is equallyeffective at decreasing sea level. Next we consider whether changes in the distribution of ocean temperaturemight be poised to cause greater basal melt to marine ice shelves.

Numerical simulations of SRM with stratospheric aerosols have shown residual SH poleward intensifiedwinds induced by differences in the vertical temperature structure of a stratospheric sulfate layer versuswell-mixed tropospheric carbon dioxide [Ammann et al., 2010; McCusker et al., 2012]. These temperatureanomalies and corresponding upper atmosphere and surface wind changes share features with those inducedby ozone depletion [Gillett and Thompson, 2003; Gillett et al., 2013; Sigmond et al., 2011; Thompson et al.,2011] and increasing greenhouse gases [Gillett et al., 2013; Sigmond et al., 2011; Polvani et al., 2011]. Althoughmechanisms and structural details differ somewhat depending on the type of upper atmospheric forcing,

MCCUSKER ET AL. STRATOSPHERIC AEROSOLS AND WAIS 4

Geophysical Research Letters 10.1002/2015GL064314

Figure 3. SH SAT and zonal wind stress anomaly maps. Annual mean surface air temperature (∘C) anomaly from 20thC(1970–1999 mean) for (a) RCP8.5, (b) Sulf, and (c) GHGrem, years 2045–2054. (d–f ) Same as Figures 3a–3c, respectively,but for zonal wind stress (N/m2). Positive indicates westerly stress on the ocean. Contours indicate sea ice extent (15%concentration contour) for the 20thC (black) and perturbed simulation (gray); note that for Figures 3e and 3f thecontours are nearly colocated. Hatching indicates statistical significance at the 90% level using the Student’s t testfor difference of two means.

the fundamental cause is the same: the zonal wind shear in the upper troposphere/lower stratosphere isincreased due to an increase in the stratospheric pole-to-equator temperature gradient. Strengthened zonalwinds then shift the jet poleward; however, what processes maintain the poleward-shifted jet is an openresearch question [Lorenz, 2014]. One study found that strengthened SH lower stratospheric/upper tropo-spheric winds increased eastward eddy propagation in the troposphere, which in turn shifts the criticallatitude for Rossby wave breaking poleward, leading to poleward shifted surface westerlies [Chen and Held,2007]. More recent studies using idealized models indicate that the poleward shift of strengthened zonalwinds is maintained by the increased equatorward reflection of poleward propagating waves at a turninglatitude, causing greater poleward momentum flux across the jet, shifting it poleward [Kidston and Vallis, 2012;Lorenz, 2014].

When aerosol concentrations are quickly increased in the stratosphere, large temperature anomalies aregenerated in the upper troposphere and stratosphere (Figure 2a). The zonal mean temperature structurewith height is a combination of strong tropical lower stratospheric heating due to absorption of tropo-spheric long-wave radiation by aerosols [Ferraro et al., 2011] and high-latitude stratospheric long-wave coolingdue to GHGs, with little change in the lower tropospheric temperature. The resultant zonal winds show astrengthened SH polar vortex and slight weakening of the subtropical jet (Figure 2b).

The upper atmospheric deviations result in surface wind and wind stress anomalies in the SH that are compa-rable to those in RCP8.5: they are poleward shifted and intensified and have similar magnitudes (Figures 3d,3e, and 4a). In contrast, the surface wind stress is very nearly returned to 20thC values in GHGrem (Figures 3fand 4a). Additionally, while both climate engineering strategies cause the sea ice edge to recover to the20thC position (20thC and climate engineering simulations show colocated 15% concentration contours inFigures 3e–3f ), annual average sea ice concentration near sea ice margins remains lower by up to 15% inSulf, but are increased beyond 20thC concentrations by almost 10% in GHGrem (Figure S1 in the supportinginformation). The residual sea ice concentration anomalies in Sulf and GHGrem are echoed in the residualtemperature anomalies shown in Figure 3: Figure 3b shows a weak local warming accompanying the reducedsea ice concentration in Sulf, while Figure 3c shows weak local cooling where there are small increases in seaice concentration in GHGrem.

4. Implications for Antarctic Ice Sheets

Both climate engineering scenarios presented here in Sulf and GHGrem greatly reduce the surface warmingover Antarctica that is seen in RCP8.5—greater than 100% in the case of GHGrem (compare Figures 3b and

MCCUSKER ET AL. STRATOSPHERIC AEROSOLS AND WAIS 5

Geophysical Research Letters 10.1002/2015GL064314

Figure 4. Southern Ocean wind stress, wind stress curl, and potential temperature. Annual mean, zonal mean (a) zonalwind stress (N/m2), and (b) curl of the wind stress (10−8 N/m3) over ocean only, as anomalies from 20thC (1970–1999mean). Positive indicates westerlies in Figure 4a and downwelling in Figure 4b. Annual mean, zonal mean oceanpotential temperature (∘C) anomalies from 20thC with depth for (c) Sulf and (d) GHGrem, years 2045–2054. Thick linesin Figures 4a and 4b indicate regions of statistical significance at the 90% level using the Student’s t test for differenceof two means. Regions encompassed by black contours in Figures 4c and 4d, which are values greater than about0.05∘C, are statistically significant at the 90% level using the Student’s t test for difference of two means.

3c with Figure 3a)—but the ocean heat uptake that occurred over the first third of the 21st century beforeclimate engineering was initiated also poses a threat to Antarctic ice sheets, particularly the WAIS, whosegrounding line is below sea level [Joughin and Alley, 2011] and retreating [Rignot et al., 2014].

Small ocean temperature perturbations can have a significant impact on basal melting of ice shelves; a warm-ing of just 0.1∘C can thin 1 m of ice shelf in a year [Rignot and Jacobs, 2002]. Zonally averaged Southern Oceantemperature anomalies in Sulf indicate that temperatures encroaching on the ice shelves average 0.15–0.20∘Cwarmer than 20thC (Figure 4c). GHGrem, however, has zonal mean temperature near the level of ice shelvesup to 0.25∘C cooler than in Sulf (Figure 4d). The ice sheet region in Antarctica most at-risk from basal meltis the Amundsen Sea Embayment sector including Pine Island Glacier (PIG). Critically, this region reveals aneven greater potential for melting under sulfate climate engineering than in the zonal average: Sulf featuresan anomalous residual warming of 0.5∘C in the 200–500 m layer (Figure 5a). The enhanced residual warm-ing in the PIG region compared to the zonal average is due to a steeper vertical temperature gradient inthe 100–400 m layer. This, combined with slightly larger vertical velocity anomalies, which will be discussednext, is the primary reason for the larger temperature anomaly in the PIG region than in the zonal aver-age. Extrapolating from Rignot and Jacobs [2002], this temperature anomaly represents a basal melt rate of5 m/yr above 20thC. In contrast, GHGrem has a cooling of the subsurface ocean up to 0.5∘C. As a point forcomparison, a recent estimate of thickness rate of change from Pine Island over the period 1994–2012 is−23 ± 3.8 m/decade, which corresponds to a volume loss of −14 ± 2 km3/yr [Paolo et al., 2015]. Hence, ourestimate of 5 m/yr (50 m/decade) from Sulf suggests that volume loss from PIG over the decade 2045–2054would be approximately twice that seen over the past two decades (however, this assumes that volume losshas a linear relation to subsurface ocean temperature, which is unlikely [Joughin et al., 2014]).

We have not taken into account the basal meltwater changes from ice shelves, because they are not resolvedin our model and the ocean response to such meltwater is uncertain. Underneath an ice shelf, rising meltwatermay drive a sub-ice shelf circulation that increases the rate CDW reaches the ice shelf base [Jacobs et al., 2011].Once meltwater exits the base of the ice shelf, it would likely rise and drive mixing near the shelf front. Whenthe meltwater reaches the surface, it may also act to strengthen the halocline and inhibit the upwelling ofheat at some distance away from the ice shelf. This would lead to the warming of waters at the depth thatmay circulate to the ice shelf base, potentially further enhancing basal melt.

GHG removal is a more effective means for cooling subsurface ocean temperatures than stratospheric aerosolinjection fundamentally because of changes in ocean circulation that are brought about by changes inatmospheric circulation. The anomalous middepth zonal average warmth evident south of about 65∘S and

MCCUSKER ET AL. STRATOSPHERIC AEROSOLS AND WAIS 6

Geophysical Research Letters 10.1002/2015GL064314

Figure 5. PIG region ocean potential temperature and vertical advection. Annual mean, zonal mean ocean potentialtemperature (∘C) with depth averaged for longitudes 80∘W to 120∘W in the Amundsen Sea Embayment/Pine IslandGlacier (PIG) region as anomalies from 20thC for (a) Sulf and (b) GHGrem. Regions encompassed by black contours,which are values greater than about 0.05∘C, are statistically significant at the 90% level using the Student’s t test fordifference of two means. The vertical dashed line shows the equatorward limit of the region used to calculate theaverages shown in Figures 5c–5e. (c) Vertical velocity averaged between 80∘W–120∘W and 65∘S–74∘S at each depthdue to Eulerian (dashed), eddy-induced (thin solid), and total motion (thick solid) as anomalies from 20thC for RCP8.5(red), Sulf (blue), and GHGrem (green) (m/yr). (d) The 20thC mean temperature gradient with depth (∘C/m). Negativeindicates increased temperature with increasing depth (z is defined positive up). (e) Same as Figure 5c but showing−Δw × dT∕dz —the largest component to the change in vertical advection (∘C/yr). Not shown are the w × dΔT∕dzcomponent, which is smaller than the Δw × dT∕dz component shown and has a pattern that does not explain theanomalous warmth in Sulf, and the Δw × dΔT∕dz component, which has a negligible magnitude compared to otherterms. Gray shading in Figures 5c and 5e indicates anomalies that are greater than 1 standard deviation of the valuesin the 1970–1999 time series of the 20thC simulation.

between 200 and 400 m depth in Sulf (Figure 4c) can be understood as anomalous upwelling of relativelywarm CDW. This upwelling is caused by increased Ekman pumping that originates from the previously dis-cussed poleward intensified zonal winds (Figures 3e) caused by stratospheric aerosols, greenhouse gases[Fyfe et al., 2007], and their combination here (Figure 4a). Engineering by sulfate aerosols features anomalousupwelling poleward of 60∘S and downwelling equatorward of 60∘S—just as in the RCP8.5 scenario (albeitwith slightly smaller upwelling than in RCP8.5). In contrast, the removal of GHGs causes a reversal of sign in thezonal wind stress and wind stress curl anomalies compared to the anomalies associated with RCP8.5, result-ing in weak anomalous downwelling south of 60∘S (Figure 4b), which results in the cooling of the 200–400 mlayer past the twentieth century mean (Figure 4d).

In the PIG region, the primary driver of the contrasting behavior between climate engineering methodsis again due to differences in circulation. Vertical velocity responses of opposing signs (Figure 5c) act on aclimatological temperature gradient that features an increase in temperature with increasing depth in theselayers (Figure 5d). The result is that anomalous vertical advection warms the water column under sulfateengineering, although it is smaller than the business-as-usual scenario, whereas removal of GHGs causesa cooling, which is primarily due to the Eulerian component of vertical advection (Figure 5e). The mid-depth residual warming resembles the “slow timescale” response to an increase in ozone forcing wherein theSouthern Ocean circulation adjusts to changes in atmospheric winds and causes ocean warming from belowthat overcomes initial SST cooling and sea ice growth on the time scale of decades [Ferreira et al., 2014].

Finally, although not germane to the stability of Antarctic ice sheets, it is interesting to note the disparateocean temperature profiles in the full zonal average equatorward of 60∘S between Sulf and GHGrem as well(Figures 4c and 4d). The warming evident equatorward of 60∘S at all depths in Sulf is a combination of residualwarmth from before climate engineering began, a poleward shift (fundamentally due to the atmosphericjet shift) of the nearly vertical climatological isotherms there, and increased downwelling of warmer surface

MCCUSKER ET AL. STRATOSPHERIC AEROSOLS AND WAIS 7

Geophysical Research Letters 10.1002/2015GL064314

waters. This feature is noticeably absent when GHGs are instead removed from the atmosphere because thejet rapidly returns to the 20thC position, resulting in weak upwelling anomalies equatorward of 60∘S. Thus,the warming equatorward of 60∘S is also due to the changes in the atmospheric winds due to the net forcingof increased GHGs and sulfate aerosols.

5. Persistent and Problematic Circulation

Circulation anomalies presented here result in a large-scale oceanic environment that would be favorablefor further thinning of Antarctic ice shelves through basal melt, particularly in the region encompassingAntarctica’s largest contributor to sea level rise [Shepherd et al., 2012], the already unstable Pine Island Glacieroutlet [Rignot et al., 2014]. Fundamentally, these ocean circulation changes are due to the modified merid-ional temperature gradient in the stratosphere and upper troposphere induced by GHGs and sulfate aerosols.As a result, the ocean circulation changes responsible for warming of the water adjacent to the ice shelvespersist for as long as GHGs and sulfates are in the atmosphere (Figures S2 and S3). Arguments that SRM withstratospheric aerosols is a “backstop” measure that could be rapidly deployed to avoid so-called climate emer-gencies, such as destabilization of marine ice sheets [Blackstock et al., 2009], are therefore not supported inthis study. Given that the PIG grounding line is currently retreating along a retrograde bed, its reversibilitymay only be achievable with substantial reductions in basal melt rate beyond present-day rates [Favier et al.,2014]—a task only removal of GHGs can accomplish based on our results.

We emphasize that surface and subsurface ocean temperatures are undoubtedly cooler when stratosphericaerosols are in use than under business-as-usual warming. However, while general circulation models suchas the one we employ do not resolve ice shelves and subscale processes that influence their thickness (e.g.,small-scale mixing, tidal currents [Joughin and Alley, 2011]), we have determined that the subsurface SouthernOcean warmth, in proximity to ice shelves that remain after sulfate aerosol deployment, is fundamentallydue to residual large-scale circulation changes. We suggest that quantitative calculations of changes in basalmelt rates, changes in ice sheet dynamics, and associated sea level rise under various climate engineeringtechniques become a high priority once coupled GCM-ice-sheet models become available. The inability tostabilize the West Antarctic ice sheet must be added to the list of weaknesses of stratospheric sulfate aerosolinjections [Robock, 2008, 2014], which already includes ozone depletion [Tilmes et al., 2008; Heckendorn et al.,2009], risk of extreme regional temperature trends upon cessation [Jones et al., 2013; McCusker et al., 2014],continued ocean acidification [Feely et al., 2004], and reduced precipitation [Bala et al., 2008].

ReferencesAmmann, C. M., W. M. Washington, G. A. Meehl, L. Buja, and H. Teng (2010), Climate engineering through artificial enhancement of natural

forcings: Magnitudes and implied consequences, J. Geophys. Res., 115, D22109, doi:10.1029/2009JD012878.Bala, G., P. B. Duffy, and K. E. Taylor (2008), Impact of geoengineering schemes on the global hydrological cycle, Proc. Natl. Acad. Sci., 105(22),

7664–7669.Blackstock, J. J., D. S. Battisti, K. Caldeira, D. M. Eardley, J. I. Katz, D. W. Keith, A. A. Patrinos, D. P. Schrag, R. H. Socolow, and S. E Koonin (2009),

Climate Engineering Responses to Climate Emergencies, Novim, Santa Barbara, Calif.Chen, G., and I. M. Held (2007), Phase speed spectra and the recent poleward shift of Southern Hemisphere surface westerlies, Geophys. Res.

Lett., 34, L21805, doi:10.1029/2007GL031200.Church, J., et al. (2013), Sea level change, in Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth

Assessment Report of the Intergovernmental Panel on Climate Change, edited by T. F. Stocker et al., Cambridge Univ. Press, Cambridge,U. K., and New York.

Church, J. A., N. J. White, and J. M. Arblaster (2005), Significant decadal-scale impact of volcanic eruptions on sea level and ocean heatcontent, Nature, 438(7064), 74–77, doi:10.1038/nature04237.

English, J. M., O. B. Toon, and M. J. Mills (2012), Microphysical simulations of sulfur burdens from stratospheric sulfur geoengineering, Atmos.Chem. Phys., 12(10), 4775–4793, doi:10.5194/acp-12-4775-2012.

Favier, L., G. Durand, S. L. Cornford, G. H. Gudmundsson, O. Gagliardini, F. Gillet-Chaulet, T. Zwinger, A. J. Payne, and A. M. Le Brocq (2014),Retreat of Pine Island Glacier controlled by marine ice-sheet instability, Nat. Clim. Change, 4, 117–121, doi:10.1038/nclimate2094.

Feely, R. A., C. L. Sabine, K. Lee, W. Berelson, J. Kleypas, V. J. Fabry, and F. J. Millero (2004), Impact of anthropogenic CO2 on the CaCO3 systemin the oceans, Science, 305(5682), 362–366, doi:10.1126/science.1097329.

Ferraro, A. J., E. J. Highwood, and A. J. Charlton-Perez (2011), Stratospheric heating by potential geoengineering aerosols, Geophys. Res. Lett.,38, L24706, doi:10.1029/2011GL049761.

Ferreira, D., J. Marshall, C. M. Bitz, S. Solomon, and A. Plumb (2014), Antarctic ocean and sea ice response to ozone depletion: Atwo-time-scale problem, J. Clim., 28, 1206–1226.

Fyfe, J. C., O. A. Saenko, K. Zickfeld, M. Eby, and A. J. Weaver (2007), The role of poleward-intensifying winds on Southern Ocean warming,J. Clim., 20(21), 5391–5400.

Gent, P. R., et al. (2011), The community climate system model version 4, J. Clim., 24(19), 4973–4991.Gillett, N. P., and D. W. J. Thompson (2003), Simulation of recent Southern Hemisphere climate change, Science, 302(5643), 273–275,

doi:10.1126/science.1087440.

AcknowledgmentsThe global climate model can beobtained from the National Center forAtmospheric Research (ncar.ucar.edu).Model simulation output data usedin the production of Figures 1–5 areavailable free of charge by emailingthe corresponding author. Thisresearch was funded by the TamakiFoundation and supported in partby the National Science Foundationthrough PLR-1341497 and TeraGridresources provided by the TexasAdvanced Computing Centerunder grant TG-ATM090059. Wethank Judy Twedt for helpfuldiscussion on experimental designand analysis and Colin Goldblatt,Alan Robock, and an anonymousreviewer for comments that improvedthe manuscript.

The Editor thanks Alan Robock andan anonymous reviewer for theirassistance in evaluating this paper.

MCCUSKER ET AL. STRATOSPHERIC AEROSOLS AND WAIS 8

Geophysical Research Letters 10.1002/2015GL064314

Gillett, N. P., V. K. Arora, K. Zickfeld, S. J. Marshall, and W. J. Merryfield (2011), Ongoing climate change following a complete cessation ofcarbon dioxide emissions, Nat. Geosci., 4(2), 83–87, doi:10.1038/ngeo1047.

Gillett, N. P., J. C. Fyfe, and D. E. Parker (2013), Attribution of observed sea level pressure trends to greenhouse gas, aerosol, and ozonechanges, Geophys. Res. Lett., 40, 2302–2306, doi:10.1002/grl.50500.

Gleckler, P. J., K. AchutaRao, J. M. Gregory, B. D. Santer, K. E. Taylor, and T. M. L. Wigley (2006), Krakatoa lives: The effect of volcanic eruptionson ocean heat content and thermal expansion, Geophys. Res. Lett., 33, L17702, doi:10.1029/2006GL026771.

Heckendorn, P., D. Weisenstein, S. Fueglistaler, B. Luo, E. Rozanov, M. Schraner, L. Thomason, and T. Peter (2009), The impact of geoengineer-ing aerosols on stratospheric temperature and ozone, Environ. Res. Lett., 4(4), 045108.

Irvine, P. J., R. L. Sriver, and K. Keller (2012), Tension between reducing sea-level rise and global warming through solar-radiation manage-ment, Nat. Clim. Change, 2(2), 97–100.

Jacobs, S. S., A. Jenkins, C. F. Giulivi, and P. Dutrieux (2011), Stronger ocean circulation and increased melting under Pine Island Glacier IceShelf, Nat. Geosci., 4(8), 519–523, doi:10.1038/ngeo1188.

Jones, A., et al. (2013), The impact of abrupt suspension of solar radiation management (termination effect) in experiment G2 of theGeoengineering Model Intercomparison Project (GeoMIP), J. Geophys. Res. Atmos., 118, 9743–9752, doi:10.1002/jgrd.50762.

Joughin, I., and R. B. Alley (2011), Stability of the West Antarctic Ice Sheet in a warming world, Nat. Geosci., 4(8), 506–513.Joughin, I., B. E. Smith, and B. Medley (2014), Marine ice sheet collapse potentially under way for the Thwaites Glacier basin, West Antarctica,

Science, 344(6185), 735–738, doi:10.1126/science.1249055.Kidston, J., and G. Vallis (2012), The relationship between the speed and the latitude of an eddy-driven jet in a stirred barotropic model,

J. Atmos. Sci., 69(11), 3251–3263.Lorenz, D. J. (2014), Understanding midlatitude jet variability and change using Rossby wave chromatography: Poleward-shifted jets in

response to external forcing, J. Atmos. Sci., 71(7), 2370–2389.McCusker, K. E., D. S. Battisti, and C. M. Bitz (2012), The climate response to stratospheric sulfate injections and implications for addressing

climate emergencies, J. Clim., 25(9), 3096–3116, doi:10.1175/JCLI-D-11-00183.1.McCusker, K. E., K. C. Armour, C. M. Bitz, and D. S. Battisti (2014), Rapid and extensive warming following cessation of solar radiation

management, Environ. Res. Lett., 9(2), 024005.Moore, J. C., S. Jevrejeva, and A. Grinsted (2010), Efficacy of geoengineering to limit 21st century sea-level rise, Proc. Natl. Acad. Sci., 107,

15,699–15,703.Notz, D. (2009), The future of ice sheets and sea ice: Between reversible retreat and unstoppable loss, Proc. Natl. Acad. Sci., 106(49),

20,590–20,595.Oppenheimer, M. (1998), Global warming and the stability of the West Antarctic Ice Sheet, Nature, 393(6683), 325–332, doi:10.1038/30661.Paolo, F. S., H. A. Fricker, and L. Padman (2015), Volume loss from Antarctic ice shelves is accelerating, Science, 348(6232), 327–331.Polvani, L. M., M. Previdi, and C. Deser (2011), Large cancellation, due to ozone recovery, of future Southern Hemisphere atmospheric

circulation trends, Geophys. Res. Lett., 38, L04707, doi:10.1029/2011GL046712.Pritchard, H. D., S. R. M. Ligtenberg, H. A. Fricker, D. G. Vaughan, M. R. van den Broeke, and L. Padman (2012), Antarctic ice-sheet loss driven

by basal melting of ice shelves, Nature, 484, 502–505, doi:10.1038/nature10968.Rasch, P. J., P. J. Crutzen, and D. B. Coleman (2008), Exploring the geoengineering of climate using stratospheric sulfate aerosols: The role of

particle size, Geophys. Res. Lett., 35, L02809, doi:10.1029/2007GL032179.Rignot, E., and S. S. Jacobs (2002), Rapid bottom melting widespread near Antarctic ice sheet grounding lines, Science, 296(5575),

2020–2023.Rignot, E., J. Mouginot, M. Morlighem, H. Seroussi, and B. Scheuchl (2014), Widespread, rapid grounding line retreat of Pine Island, Thwaites,

Smith and Kohler glaciers, West Antarctica from 1992 to 2011, Geophys. Res. Lett., 41, 3502–3509, doi:0.1002/2014GL060140.Robock, A. (2008), 20 reasons why geoengineering may be a bad idea, Bull. At. Sci., 64, 14–18.Robock, A. (2014), Stratospheric aerosol geoengineering, in Geoengineering of the Climate System, vol. 38, edited by R. M. Harrison and

R. E. Hester, pp. 162–185, Royal Soc. of Chem., London.Shepherd, A., et al. (2012), A reconciled estimate of ice-sheet mass balance, Science, 338(6111), 1183–1189, doi:10.1126/science.1228102.Sigmond, M., M. Reader, J. Fyfe, and N. Gillett (2011), Drivers of past and future Southern Ocean change: Stratospheric ozone versus

greenhouse gas impacts, Geophys. Res. Lett., 38, L12601, doi:10.1029/2011GL047120.Spence, P., S. M. Griffies, M. H. England, A. M. Hogg, O. A. Saenko, and N. C. Jourdain (2014), Rapid subsurface warming and circulation

changes of Antarctic coastal waters by poleward shifting winds, Geophys. Res. Lett., 41, 4601–4610, doi:10.1002/2014GL060613.Steig, E. J., et al. (2013), Recent climate and ice-sheet changes in West Antarctica compared with the past 2,000 years, Nat. Geosci., 6(5),

372–375, doi:10.1038/ngeo1778. [Available at http://www.nature.com/ngeo/journal/v6/n5/abs/ngeo1778.html.]Thoma, M., A. Jenkins, D. Holland, and S. Jacobs (2008), Modelling circumpolar deep water intrusions on the Amundsen Sea continental

shelf, Antarctica, Geophys. Res. Lett., 35, L18602, doi:10.1029/2008GL034939.Thompson, D. W., S. Solomon, P. J. Kushner, M. H. England, K. M. Grise, and D. J. Karoly (2011), Signatures of the Antarctic ozone hole in

Southern Hemisphere surface climate change, Nat. Geosci., 4(11), 741–749.Tilmes, S., R. Muller, and R. Salawitch (2008), The sensitivity of polar ozone depletion to proposed geoengineering schemes, Science,

320(5880), 1201–1204.Yin, J., J. T. Overpeck, S. M. Griffies, A. Hu, J. L. Russell, and R. J. Stouffer (2011), Different magnitudes of projected subsurface

ocean warming around Greenland and Antarctica, Nat. Geosci., 4(8), 524–528, doi:10.1038/ngeo1189.

MCCUSKER ET AL. STRATOSPHERIC AEROSOLS AND WAIS 9