1. Introduction
The three largest explosive volcanic eruptions to have occurred in the second half of the twentieth century were Mount Agung in Bali, Indonesia (March and May 1963), El Chichón in Chiapas, Mexico (March–April 1982), and Mount Pinatubo in the Luzon, Philippines (June 1991). These tropical eruptions emitted tens of megatons total of sulfur dioxide (SO2) and particulate material into the lower stratosphere (e.g., McInturff et al. 1971; McCormick and Veiga 1992; Baran and Foot 1994; Robock 2002) and caused a warming of the tropical lower stratosphere on the order of a few degrees Celsius (e.g., Angell and Korshover 1978; Labitzke et al. 1983; Parker and Brownscombe 1983; Angell 1993) resulting from the absorption of terrestrial longwave and solar near-infrared radiation (e.g., Newell 1970; Pollack et al. 1976; Kinne et al. 1992). Thermal fluctuations related to the tropical quasi-biennial oscillation (QBO; e.g., Angell and Korshover 1964) also contributed to the lower-stratospheric warming; however, the total warming has been shown to considerably exceed that associated with the QBO (e.g., Parker and Brownscombe 1983; Stenchikov et al. 1998). The SO2 molecules that were injected into the stratosphere oxidized within a few weeks to form sulfate aerosols that had both heating and cooling effects on the troposphere and surface below, with increased emissions of downward longwave radiation (heating) and reduced downward visible and near-infrared fluxes (cooling). The total radiative balance was affected by the distributions of clouds, water vapor, and surface temperature, which varied in response to the perturbed tropospheric climate (Stenchikov et al. 1998; Ramachandran et al. 2000), but the net effect was a subtropical and tropical cooling at the surface for about 2 yr after the eruptions (Robock and Mao 1995); however, this cooling was not observed in the first year after the 1982 El Chichón eruption because the developing 1982/83 El Niño event produced a large compensating warming (Robock 2000).
The eruptions’ impacts were global. The volcanic sulfate aerosol from the Mount Agung eruption spread mostly throughout the Southern Hemisphere, the El Chichón aerosol spread mostly throughout the Northern Hemisphere, and the Mount Pinatubo aerosol spread to both poles (Viebrock and Flowers 1968; Stenchikov et al. 1998). Both the stratosphere and troposphere responded dynamically to the perturbed radiative properties of the tropical atmosphere following the eruptions. Climate variations included a substantial winter surface warming of Northern Hemisphere continents, including over Europe and Siberia and anomalous cooling over the Middle East and Greenland, as was documented following the El Chichón and Mount Pinatubo eruptions (Robock and Mao 1992; Stenchikov et al. 1998, respectively). The most widely accepted large-scale dynamical mechanism associated with these anomalies stems from the lower-stratospheric warming in the tropics, which gives rise to an enhanced equator-to-pole temperature gradient. This produces an enhanced northern polar vortex and a response in the tropospheric circulation that gives rise to large-scale surface climate variations (e.g., Graf et al. 1993; Kodera 1994). This dynamical effect has been described as a modification of the interaction between stratospheric westerlies and vertically propagating planetary waves (Perlwitz and Graf 1995), in that an enhanced winter polar vortex leads to increased refraction of planetary waves, which in turn decreases the deceleration of the vortex (e.g., Stenchikov et al. 2006). Thus, the impact of volcanic aerosols leads to an enhanced positive phase of the Arctic Oscillation (AO) and associated North Atlantic Oscillation (NAO; Thompson and Wallace 1998) that is most prominent in boreal winter and persists for up to 2 yr after each eruption (e.g., Robock and Mao 1992; Stenchikov et al. 2006).
The observed AO response to volcanic eruptions has been reproduced with varying success in general circulation models (GCMs) over the last decade. Kirchner et al. (1999) reproduced a Northern Hemisphere winter warming pattern in the troposphere of similar strength to that observed using the ECHAM4 GCM with Mount Pinatubo aerosol parameters developed by Stenchikov et al. (1998). However, the intensity of the stratospheric polar vortex anomaly was much weaker than that observed, and its position and strength were strongly influenced by sea surface temperature (SST) forcing. This made it difficult to assess the effect of the stratosphere on the tropospheric circulation, and internal variability likely played a role in the five-member ensemble mean surface temperature response. Further, their model did not account for the QBO, which is known to have a significant effect on the winter circulation in the Northern Hemisphere extratropics (e.g., Holton and Tan 1980, 1982; Dunkerton and Baldwin 1991). Ramachandran et al. (2000) used the Geophysical Fluid Dynamics Laboratory (GFDL) “SKYHI” GCM with Mount Pinatubo aerosol parameters similar to those developed by Stenchikov et al. (1998). They generated substantial longwave stratospheric cooling in the middle to high latitudes that contributed to the enhanced equator-to-pole temperature gradient initiated by the tropical lower-stratospheric warming. The impact on the polar vortex was not assessed, however, because of the relatively high interannual variability and low signal-to-noise ratio in high-latitude stratospheric temperature anomalies compared to those observed. Further, their model also neglected the QBO. Neither the Stenchikov et al. (1998) nor Andronova et al. (1999) modeling studies generated strong extratropical longwave cooling following the Mount Pinatubo eruption; Ramachandran et al. (2000) attributed this cooling effect to the impact of high clouds in increasing the thermal emission by the aerosol layer.
Stenchikov et al. (2002) examined the sensitivity of the GFDL SKYHI GCM to short-term ozone loss after the Mount Pinatubo eruption, which is caused by heterogeneous and radiative volcanic aerosol effects on ozone photochemistry (e.g., Hofmann and Solomon 1989; Solomon et al. 1998). Experiments that used post–Mount Pinatubo ozone anomalies produced a strong positive phase of the AO comparable to that observed, and a weaker positive AO response was produced by experiments that included only the tropospheric effects of aerosols (i.e., no stratospheric heating); however, interpretation of the results was again limited by the absence of a QBO in the model. The importance of the QBO was subsequently demonstrated by Stenchikov et al. (2004) in simulations that included post–Mount Pinatubo volcanic aerosols and a realistic QBO simulation, but without volcanically induced ozone depletion. The winter response to the combined effect of aerosols and the QBO was a positive AO phase that was enhanced by the westerly QBO phase during the 1992/93 winter, as observed.
More recently Jones et al. (2005) demonstrated the potential impact of a strengthened equator-to-pole stratospheric temperature gradient on the winter vortex following a generic volcanic “supereruption” (on the scale of 100 times larger than that of Mount Pinatubo), the last of which was the Toba eruption in Sumatra, Indonesia (around 71,000–75,000 years ago; see Rose and Chesner 1990; Zielinski et al. 1996; Oppenheimer 2002). Tropical lower-stratospheric warming from volcanic aerosol prescribed in their experiment increased the meridional temperature gradient, which led to a strong positive phase of the AO and subsequent winter warming over northern landmasses in the first two winters following the eruption. The winter warming was attributed in equal parts to the positive AO mode (which dominated the response south of the Arctic Circle) and increased longwave forcing from the stratospheric aerosols (which dominated north of the circle). The high-latitude response to nineteenth- and twentieth-century volcanic eruptions simulated in many present-day GCMs remains poor, however. Stenchikov et al. (2006) assessed the response to nine volcanic eruptions during the period 1860–1999 in long climate simulations using seven climate models included in the model intercomparison conducted as part of the Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4; Alley et al. 2007). The models tended to simulate a positive AO phase in response to volcanic forcing, however this was much weaker than that observed. Stenchikov et al. (2006) noted the coarseness of the models’ resolution and suggested that an improved response may be achieved by both increasing the vertical resolution in the stratosphere and extending the model lid into the mesosphere.
This study uses two versions of the Met Office Hadley Centre climate model with different vertical resolutions above the tropopause to investigate the extratropical winter response to volcanic aerosol forcing and its sensitivity to stratospheric resolution in seasonal hindcasts. Model experiments and observational data are described in section 2, results are presented in section 3, and conclusions are summarized in section 4.
2. Data description
a. Model experiments
We make use of the 38- and 60-level atmosphere-only configurations of the first-generation Hadley Centre Global Environmental Model (HadGEM1; Martin et al. 2006), with many of the changes proposed for HadGEM2-A as documented in Collins et al. (2008), and a spatial resolution of 1.25° latitude × 1.875° longitude. The 38-level version (L38) has a model top in the stratosphere at 39.3 km (∼5 mb) and the 60-level version (L60) has a model top near the mesopause at 84.1 km (∼0.005 mb). Both models have the same vertical resolution in the troposphere, a horizontal resolution of N96, and a time step of 20 min, allowing for a clean assessment of the impact of volcanic aerosol forcing. All model experiments described here are forced with time-varying greenhouse gas concentrations, including CO2, CH4, N2O, CFCl3, and CF2Cl2, and changes in vegetation, sulfur, soot, and biomass emissions. Sea surface temperature and sea ice extent variations are specified from an analysis of historical observations (Rayner et al. 2003), and atmospheric ozone concentrations are held constant at 1990 levels.
Three experiments are devised for the L38 and L60 models. The first of these incorporates stratospheric volcanic aerosol column mass into winter [December–February (DJF)] hindcasts for 1963/64, 1964/65, 1982/83, 1983/84, 1991/92, and 1992/93. As discussed, the main atmospheric thermal and dynamical effects of recent volcanic eruptions have been shown to persist for about 2 yr after each eruption, and thus we produce L38 and L60 model hindcasts for the two boreal winters following the three largest explosive volcanic eruptions that have occurred over the last 50 yr: Mount Agung (March–May 1963), El Chichón (March–April 1982), and Mount Pinatubo (June 1991). Fifteen-member ensembles are produced for each of these six “postvolcanic” winter periods. The atmospheric conditions for each ensemble are initialized using 6-hourly data from the 40-yr European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40; Uppala et al. 2005), starting from 1200 UTC 27 November and ending at 0000 UTC 1 December, to produce 15 hindcasts that are integrated from a model start date of 1 December.
Stratospheric volcanic aerosol column mass totals are calculated from the updated monthly optical depth dataset of Sato et al. (1993; updates are documented online at http://www.giss.nasa.gov/data/strataer/). The aerosol is applied globally to the model across four latitude bands of approximately equal areas (90°–30°S, 30°S–0°, 0°–30°N, and 30°–90°N) and above the tropopause as a simple step function with an equal mass mixing ratio (MMR) at all altitudes; thus, the aerosol mass density decreases with altitude. The ejected Mount Pinatubo volcanic aerosol was confined to the lower- and midstratosphere with the bulk of the aerosol cloud below 30 mb (Stenchikov et al. 1998; Andronova et al. 1999), and hence the aerosol is applied in our experiments from the tropopause to 30 mb. The radiative effect of aerosols is modeled by scattering and absorbing incoming solar radiation; most of the scattering is at the shorter wavelengths and most of the absorption is at the longer wavelengths including near-infrared. The aerosol also absorbs longwave radiation emitted by the surface and troposphere. Factors such as volcanic ash, dust, smoke, and changes in tropospheric aerosols and their cloud–aerosol interactions are not incorporated because only the effects of changes in stratospheric aerosols are included (Jones et al. 2005). The time evolution of aerosol distributed across the four latitude bands is shown in Fig. 1 for each volcanic eruption.
The second experiment provides a “nonvolcanic” winter climatology for calculating extratropical anomalies in response to volcanic forcing. We thus produce 5-month model hindcasts for a subset of years during which no volcanic eruptions occurred: 1968/69, 1987/88, 1989/90, 1995/96, and 1997/98. Note that there is no net AO signal averaged over these winters. These years are also chosen so that composite anomalies calculated relative to climatology are not biased by cold/warm El Niño–Southern Oscillation (ENSO) episodes, easterly/westerly QBO phases, or stratospheric sudden warming events, which may otherwise contaminate the anomalous postvolcanic signal in the extratropics (assuming a linear superposition) by modulating the northern winter polar vortex (e.g., Hamilton 1993; Baldwin et al. 2001). Fifteen-member ensembles for each of these nonvolcanic winter periods are also initialized at 6-hourly intervals starting from 1200 UTC 27 November and ending with 0000 UTC 1 December to produce hindcasts that are integrated from a model start date of 1 December.
The third experiment, referred to as “NOVOLC,” is a repeat of the first experiment (i.e., for the six postvolcanic winters), but without volcanic aerosol forcing in the models. This additional set of hindcasts allows us to separate the impact of volcanic aerosol from the role of initial/boundary condition forcing in the hindcasts (section 3c).
b. Observational data
We use the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis 1 dataset (NNR1; Kalnay et al. 1996) for model comparison and assessment. NNR1 is a global reanalysis spanning the period 1948–2005, produced with a model resolution of T62L28, and provides global, quality-controlled datasets using a “frozen” data assimilation–forecast system. The use of such a system prevents problems of pseudoclimate signals being introduced into the dataset through changes in assimilation techniques and model formulation.
To isolate the observed volcanic signal we chose nonvolcanic winter reference periods between 1956 and 1991 during which no volcanic eruptions occurred and the atmosphere was relatively clear of volcanic aerosol [such that global mean monthly optical thicknesses at 550 nm were less than 0.015 (from Sato et al. 1993)]. These DJF periods fell within December 1957–February 1961, December 1971–February 1974, December 1976–February 1981, and December 1985–February 1991. With the exception of the earliest four winters, these periods overlap with the recent nonvolcanic reference periods used by Stenchikov et al. (2006). Observed volcanic anomaly composites are thus calculated in a manner similar to that used for the ensemble mean composites as postvolcanic minus nonvolcanic winter averages, and are also not biased by cold/warm ENSO episodes, easterly/westerly QBO phases, or stratospheric sudden warming events. Note that fewer years are used in the model climatology (5) than are used in the observed climatology (18) because of the limitations imposed by computational expense. However with 15 hindcasts per ensemble, the total number of realizations in the model climatology exceeds the total number of winters used in the observed climatology and provides a high signal-to-noise ratio in our analyses.
3. Model response to volcanic aerosol forcing
a. Robustness of observed signal
The strength of the observed extratropical response to volcanic forcing is not well represented in present-day climate simulations despite the models’ ability to reproduce radiative warming of the lower tropical stratosphere (e.g., Stenchikov et al. 2006). This suggests possible model deficiencies in capturing the subsequent dynamical response that affects high-latitude circulation anomalies; however, there may be an alternative explanation. It is possible that the observed signal is an artifact of internal variability rather than a volcanic forced signal. If this were the case, it would be possible to readily reproduce the observed lower-stratospheric AO signal in a model without volcanic forcing using a single realization for each of the six winters. We test this here by randomly sampling single-model realizations over the six winters to produce composites of six realizations from which we exclude the forced response to volcanic eruptions. We repeat this for all possible single-realization combinations to generate 156 winter composites of anomalous geopotential height at 50 mb (Z50) for both the L38 and L60 models. This is quantified and presented in a meaningful way using a bar chart metric for which Z50 anomalies averaged poleward of 65°N are calculated from each composite and assigned to frequency bins. The number of six winter samples (expressed as a percentage of the total) that reproduce or exceed the strength of the observed response is used to assess the likelihood of the observed signal being due to internal variability, for which we adopt the parlance used by Alley et al. (2007): The strong observed AO signal is extremely unlikely to be an artifact of internal variability (or extremely likely to be a robust response to volcanic forcing) if it is reproduced in less than 5% of the model iterations.
The results are presented in Fig. 2 for the L38 and L60 models. Note that the integrated Z50 anomaly over all bins gives an average polar Z50 anomaly near zero. The percentage of iterations that reproduce or exceed the strength of the observed response for each metric are shown in parentheses in the title for each plot. The observed strength of winter Z50 anomalies is reproduced in only 0.9% and 2.2% of iterations for L38 and L60, respectively, and it is therefore extremely unlikely that the observed signal is due to internal variability. We further show that HadGEM1 reproduces realistic AO variability by conducting a covariance EOF analysis of Z50 for experiments without volcanic forcing; the leading EOF for the models is similar to that for NNR1, with a peak magnitude around 140–160 m (Fig. 3). We can thus conclude with more than 95% confidence that climate models are failing to capture a robust observed high-latitude response to volcanic aerosol.
b. Extratropical response to volcanic aerosol forcing
The observed AO response to the three volcanic eruptions is illustrated in Fig. 4. A significant strengthening of the winter polar vortex is seen in the NNR1 analysis, with Z50 anomalies reaching −240 m near the North Pole. This represents an anomalously positive phase of the AO and is characteristic of the stratospheric response to low-latitude volcanic eruptions observed over the last 150 yr (Stenchikov et al. 2006). The surface signal is characterized by anomalous warming of up to 2 K over northern Europe (north of 50°N in Fig. 4) and anomalous cooling that reaches −1 K over southern Europe (south of 50°N). Figure 4 shows that the L38 and L60 models capture a statistically significant strengthening of the polar vortex with a peak Z50 magnitude around 70%–75% of that observed and centered just off the pole. The observed surface signal is also broadly reproduced by each model with statistically significant anomalies over both northern and southern Europe. The positive AO signal in the models is further seen in 50- and 850-mb zonal wind, mean sea level pressure, and surface precipitation anomalies for both models (not shown), highlighting the fact that the models adequately reproduce the observed signal. We also note that both model configurations give a similar response: differences in the strength of the European AO signal between the models are not statistically significant. We therefore conclude that increasing the vertical resolution in the stratosphere and extending the model lid into the mesosphere does not significantly affect the AO/NAO response to volcanic aerosol forcing in our seasonal hindcasts.
c. Extratropical response without volcanic aerosol forcing
The winter response of our initialized hindcasts to volcanic forcing is not consistent with the IPCC AR4 climate models evaluated by Stenchikov et al. (2006), which tended to produce only a weak positive AO response. A major difference here is that our model is run in seasonal hindcast mode while those in recent studies were run as long climate simulations. We thus consider the possibility that the AO signal captured by our models may in fact originate from initial conditions and surface boundary forcing (SST) anomalies that contain the observed signal, rather than from the implementation of volcanic aerosol itself. We test this by repeating the analysis of the previous section except for the NOVOLC experiments relative to climatology. These experiments therefore only include the influence of initial and boundary condition forcing on the extratropical winter circulation, with the absence of volcanic aerosol. Surprisingly, the stratospheric polar and surface European anomalies in Fig. 5 are similar to those shown in Fig. 4, revealing that the strong intensification of the polar vortex in the model hindcasts is in fact due to either persistence of the observed signal present in the initial conditions or forcing from observed SSTs. The two figures only differ near the tropics where the presence of volcanic aerosol forcing in the lower stratosphere leads to an anomalous Z50 increase resulting from enhanced radiative warming.
For completeness we present the difference between Figs. 4 and 5; a composite of postvolcanic experiments relative to the NOVOLC experiments to see the impact of volcanic aerosol forcing alone, without the influence of initial/boundary condition forcing (since the initial and boundary conditions are identical for postvolcanic and NOVOLC experiments). The weak stratospheric polar and surface European signals seen in Fig. 6 confirm that, consistent with the IPCC AR4 climate simulations in Stenchikov et al. (2006), our model hindcasts fail to adequately capture the observed extratropical winter response to volcanic aerosol forcing alone. This is despite the fact that the models demonstrate a realistic lower-stratospheric signal associated with aerosol heating at low latitudes (Figs. 4 and 6).
d. Impact of initial and boundary condition forcing
We finally attempt to further understand the impact of initial and boundary condition forcing on the extratropical winter signal in the models by examining the December–February signals separately. Figure 7 presents monthly mean Z50 anomalies in NNR1 and each model, composited over all postvolcanic years (relative to climatology), with error bars representing 90% confidence intervals based on model ensemble spread using a two-tailed t test. The anomaly for December in each model is near −200 m, similar to that observed. If this signal was being forced into the model through the surface (SST) boundary we would expect it to persist over the entire winter. However, the strength of the model anomaly is halved in January and the signal is completely gone in February, while the observed response remains strong through the winter. (We note that small differences between the observed and modeled monthly anomalies may arise from the larger number of years used in the observed climatology.) This result strongly suggests the role of initial conditions in forcing the model with the observed extratropical response to volcanic eruptions at the beginning of the hindcast period. This is further supported by an atmosphere-only long climate simulation using the L60 model, which also shows a weak postvolcanic extratropical signal despite being forced with both volcanic aerosol and observed SSTs (not shown).
The persistence of the AO signal in the first month of the hindcasts is consistent with the 30–35-day radiative time scale in the lower stratosphere of the northern polar region during winter (Kiehl and Solomon 1986; Newman and Rosenfield 1997). Moreover, previous studies have shown that, while statistical forecasts of northern European winter surface anomalies decrease in skill with increasing lead time, there is an increase in skill at 30–35 days when the forecasts include stratospheric information (e.g., Christiansen 2005). The results of our study thus lend support to the premise that seasonal winter forecasts are substantially improved with the inclusion of lower-stratospheric anomalies (Charlton et al. 2003; Christiansen 2005).
4. Summary, discussion, and perspective
The extratropical winter response to volcanic aerosol forcing and its sensitivity to stratospheric resolution in seasonal hindcasts are investigated in two versions of the Met Office Hadley Centre climate model with different vertical resolutions above the tropopause. Seasonal hindcasts initialized on 1 December produce a strengthening of the winter polar vortex and anomalous warming over northern Europe characteristic of the positive phase of the Arctic Oscillation (AO) when forced with volcanic aerosol following the 1963 Mount Agung, 1982 El Chichón, and 1991 Mount Pinatubo eruptions, as is observed. The AO signal in the L60 extended model is of comparable strength to that in the L38 standard model, suggesting that there is little impact from both increasing the vertical resolution in the stratosphere and extending the model domain to near the mesopause. The presence of this signal in the models, however, is likely due to the persistence of the observed signal from the initial conditions, with the models showing a weak response to volcanic aerosol forcing alone that is consistent with previous studies on the impact of volcanic forcing in long climate simulations (e.g., Stenchikov et al. 2006). The persistence of the observed signal in the first month of the hindcasts is consistent with the 30–35-day radiative time scale in the lower stratosphere and supports the premise that seasonal winter forecasts are substantially improved with the inclusion of stratospheric information. Hence, the models are able to simulate the response to volcanic eruptions if initialized in early winter, but they do not demonstrate extratropical circulation sensitivity to volcanic aerosol forcing.
It is presently not clear why climate models show a weak extratropical response to volcanic aerosol. In the early stages of this work we examined the possibility that the polar stratosphere is sensitive to changes in the vertical distribution of volcanic aerosol in the tropical stratosphere (not shown). Postvolcanic sensitivity experiments (relative to NOVOLC experiments) showed that while the models reasonably simulated associated changes in lower-stratospheric tropical heating, the extratropical winter response remained weak, suggesting a failure to generate a strong feedback between the enhanced equator-to-pole temperature gradient and circulation anomalies at high latitudes. Future work to untangle the possible reasons behind these model deficiencies could include (i) implementing a more realistic spatial structure of volcanic aerosol that varies smoothly in latitude, longitude, and time (particularly in the vicinity of the polar night jet); and (ii) taking account of the volcanically induced ozone depletion that was observed in the midlatitudes following the El Chichón and Mount Pinatubo eruptions (e.g., Hofmann and Solomon 1989; Angell 1997; Solomon 1999), which may intensify the polar vortex through high-latitude radiative cooling (Stenchikov et al. 2002). Note however that the observed AO response to the nine largest volcanic eruptions from 1860 to 1999 [of which 7 occurred prior to 1980 (Stenchikov et al. 2006)] is of similar strength to the observed response to the two post-1980 eruptions alone. The possible role of volcanically induced ozone depletion observed after 1980 in strengthening the AO response to volcanic forcing is therefore unlikely to explain the full discrepancy between models and observations.
Despite the model deficiencies seen here, HadGEM1 produces realistic AO variability (Fig. 3) and a realistic AO response to ENSO, the QBO, and stratospheric sudden warming events (Marshall and Scaife 2009; Marshall and Scaife 2009, manuscript submitted to J. Geophys. Res.). The model’s failure to respond to volcanic aerosol forcing, however, may suggest an issue regarding the high-latitude climatology of the model; specifically, the propagation of planetary waves from the troposphere into the polar stratosphere, which can influence the high-latitude dynamical response to volcanic forcing (Kodera 1994; Perlwitz and Harnik 2003). Characteristics of wave propagation will be influenced by subtle differences between observed and modeled mean climates or by a model bias in the generation of planetary waves, such that the model may be overrepresenting wave propagation into the polar stratosphere leading to a destabilization of the postvolcanic winter vortex. Potential biases in the planetary wave response are supported by an out-of-phase relationship between the modeled and observed standing waves, as is suggested by the nonzonal component of the response seen in Fig. 4. The likely importance of planetary wave refraction is also supported by the absence of significant winter anomalies in the observed annular mode of the Southern Hemisphere where planetary wave activity is weak and the strong polar vortex remains resistant to perturbations forced by volcanic aerosols (Robock et al. 2007). Investigation of this issue also requires further detailed analysis.
Acknowledgments
The authors thank Gareth Jones at the Met Office Hadley Centre, and Manoj Joshi and Keith Shine at the University of Reading for helpful discussions during the course of this work. We also thank the three anonymous reviewers for suggested improvements to the manuscript. ERA-40 data used in this study were provided by ECMWF, and the NCEP–NCAR reanalyses were provided by the NOAA/CIRES Climate Diagnostics Center. This work was supported by the Joint DECC and MoD Integrated Climate Programme (DECC) GA01101, (MoD) CBC/2B/0417_Annex C5.
REFERENCES
Alley, R. B., and Coauthors, 2007: Summary for policymakers. Climate Change 2007: The Physical Science Basis, S. Solomon et al., Eds., Cambridge University Press, 18 pp.
Andronova, N. G., E. V. Rozanov, F. Yang, M. E. Schlesinger, and G. L. Stenchikov, 1999: Radiative forcing by volcanic aerosols from 1850 to 1994. J. Geophys. Res., 104 , 16807–16826.
Angell, J. K., 1993: Comparison of stratospheric warming following Agung, El Chichón and Pinatubo volcanic eruptions. Geophys. Res. Lett., 20 , 715–718.
Angell, J. K., 1997: Estimated impact of Agung, El Chichón and Pinatubo volcanic eruptions on global and regional total ozone after adjustment for the QBO. Geophys. Res. Lett., 24 , 647–650.
Angell, J. K., and J. Korshover, 1964: Quasi-biennial variations in temperature, total ozone, and tropopause height. J. Atmos. Sci., 21 , 479–492.
Angell, J. K., and J. Korshover, 1978: Estimate of global temperature variations in the 100–30 mb layer between 1958 and 1977. Mon. Wea. Rev., 106 , 1422–1432.
Baldwin, M. P., and Coauthors, 2001: The quasi-biennial oscillation. Rev. Geophys., 39 , 179–229.
Baran, A. J., and J. S. Foot, 1994: A new application of the operational sounder HIRS in determining a climatology of sulphuric acid aerosol from the Pinatubo eruption. J. Geophys. Res., 99 , 25673–25679.
Charlton, A. J., A. O’Neill, D. B. Stephenson, W. A. Lahoz, and M. P. Baldwin, 2003: Can knowledge of the state of the stratosphere be used to improve statistical forecasts of the troposphere? Quart. J. Roy. Meteor. Soc., 129 , 3205–3224.
Christiansen, B., 2005: Downward propagation and statistical forecast of the near-surface weather. J. Geophys. Res., 110 , D14104. doi:10.1029/2004JD005431.
Collins, W. J., and Coauthors, 2008: Evaluation of the HadGEM2 model. Met Office Hadley Centre Tech. Note No. 74, 47 pp. [Available online at http://www.metoffice.gov.uk/publications/HCTN/HCTN_74.pdf].
Dunkerton, T. J., and M. P. Baldwin, 1991: Quasi-biennial modulation of planetary wave fluxes in the Northern Hemisphere winter. J. Atmos. Sci., 48 , 1041–1043.
Graf, H-F., I. Kirchner, A. Robock, and I. Schult, 1993: Pinatubo eruption winter climate effects: Model versus observations. Climate Dyn., 9 , 81–93.
Hamilton, K., 1993: An examination of observed Southern Oscillation effects in the Northern Hemisphere stratosphere. J. Atmos. Sci., 50 , 3468–3473.
Hofmann, D. J., and S. Solomon, 1989: Ozone destruction through heterogeneous chemistry following the eruption of El Chichón. J. Geophys. Res., 94 , 5029–5041.
Holton, J. R., and H. C. Tan, 1980: The influence of the equatorial quasi-biennial oscillation on the global circulation at 50 mb. J. Atmos. Sci., 37 , 2200–2208.
Holton, J. R., and H. C. Tan, 1982: The quasi-biennial oscillation in the Northern Hemisphere lower stratosphere. J. Meteor. Soc. Japan, 60 , 140–147.
Jones, G. S., J. M. Gregory, P. A. Stott, S. F. B. Tett, and R. B. Thorpe, 2005: An AOGCM simulation of the climate response to a volcanic super-eruption. Climate Dyn., 25 , 725–738. doi:10.1007/s00382-005-0066-8.
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77 , 437–471.
Kiehl, J. T., and S. Solomon, 1986: On the radiative balance of the stratosphere. J. Atmos. Sci., 43 , 1525–1534.
Kinne, S., O. B. Toon, and M. J. Prather, 1992: Buffering of stratospheric circulation by changing amounts of tropical ozone, a Pinatubo case study. Geophys. Res. Lett., 19 , 1927–1930.
Kirchner, I., G. L. Stenchikov, H-F. Graf, A. Robock, and J. C. Antuña, 1999: Climate model simulation of winter warming and summer cooling following the 1991 Mount Pinatubo volcanic eruption. J. Geophys. Res., 104 , 19039–19055.
Kodera, K., 1994: Influence of volcanic eruptions on the troposphere through stratospheric dynamical processes in the Northern Hemisphere winter. J. Geophys. Res., 99 , 1273–1282.
Labitzke, K., B. Naujokat, and M. P. McCormick, 1983: Temperature effects on the stratosphere of the Arpil 4, 1982 eruption of El Chichón, Mexico. Geophys. Res. Lett., 10 , 24–26.
Marshall, A. G., and A. A. Scaife, 2009: Impact of the QBO on surface winter climate. J. Geophys. Res., 114 , D18110. doi:10.1029/2009JD011737.
Martin, G. M., M. A. Ringer, V. D. Pope, A. Jones, C. Dearden, and T. J. Hinton, 2006: The physical properties of the atmosphere in the new Hadley Centre Global Environmental Model (HadGEM1). Part I: Model description and global climatology. J. Climate, 19 , 1274–1301.
McCormick, M. P., and R. E. Veiga, 1992: SAGE II measurements of early Pinatubo aerosols. Geophys. Res. Lett., 19 , 155–158.
McInturff, R. M., A. J. Miller, J. K. Angell, and J. Korshover, 1971: Possible effects on the stratosphere of the 1963 Mt. Agung volcanic eruption. J. Atmos. Sci., 28 , 1304–1307.
Newell, R. E., 1970: Stratospheric temperature change from the Mt. Agung volcanic eruption of 1963. J. Atmos. Sci., 27 , 977–978.
Newman, P. A., and J. E. Rosenfield, 1997: Stratospheric thermal damping times. Geophys. Res. Lett., 24 , 433–436.
Oppenheimer, C., 2002: Limited global change due to the largest known Quaternary eruption, Toba ≈ 74 kyr bp? Quat. Sci. Rev., 21 , 1593–1609.
Parker, D. E., and J. L. Brownscombe, 1983: Stratospheric warming following the El Chichón volcanic eruption. Nature, 310 , 406–408.
Perlwitz, J., and H-F. Graf, 1995: The statistical connection between tropospheric and stratospheric circulation of the Northern Hemisphere in winter. J. Climate, 8 , 2281–2295.
Perlwitz, J., and N. Harnik, 2003: Observational evidence of a stratospheric influence on the troposphere by planetary wave reflection. J. Climate, 16 , 3011–3026.
Pollack, J. B., O. B. Toon, C. Sagan, A. Summers, B. Baldwin, and W. Van Camp, 1976: Volcanic explosions and climatic change: A theoretical assessment. J. Geophys. Res., 81 , 1071–1083.
Ramachandran, S., V. Ramaswamy, G. L. Stenchikov, and A. Robock, 2000: Radiative impact of the Mount Pinatubo volcanic eruption: Lower stratospheric response. J. Geophys. Res., 105 , 24409–24429.
Rayner, N. A., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan, 2003: Globally complete analyses of sea surface temperature, sea ice and night marine air temperature, 1871-2000. J. Geophys. Res., 108 , 4407. doi:10.1029/2002JD002670.
Robock, A., 2000: Volcanic eruptions and climate. Rev. Geophys., 38 , 191–219.
Robock, A., 2002: The Earth system: Physical and chemical dimensions of global environmental change. Encyclopedia of Global Environmental Change, Vol. 1, M. C. MacCracken and J. S. Perry, Eds., Wiley, 736 pp.
Robock, A., and J. Mao, 1992: Winter warming from large volcanic eruptions. Geophys. Res. Lett., 19 , 2405–2408.
Robock, A., and J. Mao, 1995: The volcanic signal in surface temperature observations. J. Climate, 8 , 1086–1103.
Robock, A., T. Adams, M. Moore, L. Oman, and G. Stenchikov, 2007: Southern Hemisphere atmospheric circulation effects of the 1991 Mount Pinatubo eruption. Geophys. Res. Lett., 34 , L23710. doi:10.1029/2007GL031403.
Rose, W., and C. Chesner, 1990: Worldwide dispersal of ash and gases from Earth’s largest known eruption: Toba, Sumatra, 75 ka. Global Planet. Change, 89 , 269–275.
Sato, M., J. E. Hansen, M. P. McCormick, and J. B. Pollack, 1993: Stratospheric aerosol optical depth, 1850-1990. J. Geophys. Res., 98 , 22987–22994.
Solomon, S., 1999: Stratospheric ozone depletion: A review of concepts and history. Rev. Geophys., 37 , 275–316.
Solomon, S., and Coauthors, 1998: Ozone depletion at mid-latitudes: Coupling of volcanic aerosols and temperature variability to anthropogenic chlorine. Geophys. Res. Lett., 25 , 1871–1874.
Spiegel, M. R., 1988: Schaum’s Outline of Theory and Problems of Statistics. 2nd ed. McGraw-Hill, 460 pp.
Stenchikov, G. L., I. Kirchner, A. Robock, H-F. Graf, J. C. Antuña, R. G. Grainger, A. Lambert, and L. Thomason, 1998: Radiative forcing from the 1991 Mount Pinatubo volcanic eruption. J. Geophys. Res., 103 , 13837–13857.
Stenchikov, G. L., A. Robock, V. Ramaswamy, M. D. Schwarzkopf, K. Hamilton, and S. Ramachandran, 2002: Arctic Oscillation response to the 1991 Mount Pinatubo eruption: Effects of volcanic aerosols and ozone depletion. J. Geophys. Res., 107 , 4803. doi:10.1029/2002JD002090.
Stenchikov, G. L., K. Hamilton, A. Robock, V. Ramaswamy, and M. D. Schwarzkopf, 2004: Arctic Oscillation response to the 1991 Pinatubo eruption in the SKYHI general circulation model with a realistic quasi-biennial oscillation. J. Geophys. Res., 109 , D03112. doi:10.1029/2003JD003699.
Stenchikov, G. L., K. Hamilton, R. J. Stouffer, A. Robock, V. Ramaswamy, B. Santer, and H-F. Graf, 2006: Climate impacts of volcanic eruptions in the IPCC AR4 climate models. J. Geophys. Res., 111 , D07107. doi:10.1029/2005JD006286.
Thompson, D. W. J., and J. M. Wallace, 1998: The Arctic Oscillation signature in the wintertime geopotential height and temperature fields. Geophys. Res. Lett., 25 , 1297–1300.
Uppala, S. M., and Coauthors, 2005: The ERA-40 re-analysis. Quart. J. Roy. Meteor. Soc., 131 , 2961–3012.
Viebrock, H. J., and E. C. Flowers, 1968: Comments on the recent decrease in solar radiation at the South Pole. Tellus, 20 , 400–411.
Zielinski, G., P. Mayewski, L. Meeker, S. Whitlow, and M. Twickler, 1996: Potential atmospheric impact of the Toba mega-eruption ∼71,000 years ago. Geophys. Res. Lett., 23 , 837–840.
Time evolution of volcanic stratospheric aerosol applied globally to the model across four latitude bands for each eruption. The first major explosive eruption for each event occurs during month 2. Latitude bands have approximately equal area: 1) 30°–90°N, 2) 0°–30°N, 3) 30° S–0°, and 4) 90°–30°S. Shading corresponds to stratospheric aerosol optical depth values at 0.55 μm of 0–0.05 (white), 0.05–0.10 (light gray), 0.10–0.15 (dark gray), and 0.15–0.20 (black).
Citation: Journal of Climate 22, 23; 10.1175/2009JCLI3145.1
Number of six-member composites (y) that pertain to each anomaly bin (x) for randomly sampling single model realizations, without the forced AO response to volcanic eruptions, over six winters. A total of 156 winter (DJF) composites of anomalous 50-mb geopotential height averaged poleward of 65°N are sampled for L38 and L60. Values lower than the observed anomaly are shaded gray, and the number of composites (expressed as a percentage of the total) that reproduce or exceed the strength of the observed response to volcanic aerosol is shown in parentheses in the title for each plot. Every third tick mark on the x axis is labeled, with a bin width of 20 m.
Citation: Journal of Climate 22, 23; 10.1175/2009JCLI3145.1
Leading covariance EOF of winter (DJF) 50-mb geopotential height for NNR1 data, and also for L38 and L60 NOVOLC experiments. Contour interval is 20 m, with gray areas indicating negative anomalies and dark gray areas indicating anomalies less than −100 m.
Citation: Journal of Climate 22, 23; 10.1175/2009JCLI3145.1
The 50-mb (left) geopotential height (Z50) and (right) surface temperature anomalies for winter (DJF) averaged over six winters with volcanic forcing relative to climatology for NNR1, L38, and L60. Stippling indicates statistically significant anomalies at the 90% confidence level using a two-tailed t test (Spiegel 1988). For Z50, the contour interval is 20 m with gray areas indicating negative anomalies and dark gray areas indicating anomalies less than −100 m. For temperature, contour interval is 0.5 K with gray areas indicating positive anomalies and dark gray areas indicating anomalies greater than 1 K.
Citation: Journal of Climate 22, 23; 10.1175/2009JCLI3145.1
The 50-mb (left) geopotential height (Z50) and (right) surface temperature anomalies for winter (DJF) averaged over six winters for NOVOLC experiments relative to climatology for L38 and L60. Stippling indicates statistically significant anomalies at the 90% confidence level using a two-tailed t test (Spiegel 1988). For Z50, the contour interval is 20 m, with gray areas indicating negative anomalies and dark gray areas indicating anomalies less than −100 m. For temperature, contour interval is 0.5 K with gray areas indicating positive anomalies and dark gray areas indicating anomalies greater than 1 K.
Citation: Journal of Climate 22, 23; 10.1175/2009JCLI3145.1
The 50-mb (left) geopotential height (Z50) and (right) surface temperature anomalies for winter (DJF) averaged over six winters with volcanic forcing relative to NOVOLC experiments for L38 and L60. Stippling indicates statistically significant anomalies at the 90% confidence level using a two-tailed t test (Spiegel 1988). For Z50, the contour interval is 20 m with gray areas indicating negative anomalies and dark gray areas indicating anomalies less than −100 m. For temperature, contour interval is 0.5 K with gray areas indicating positive anomalies and dark gray areas indicating anomalies greater than 1 K.
Citation: Journal of Climate 22, 23; 10.1175/2009JCLI3145.1
Monthly mean Z50 anomalies for December–February averaged over six winters with volcanic forcing relative to climatology for NNR1, L38, and L60. Error bars represent 90% confidence intervals based on model ensemble spread using a two-tailed t test (Spiegel 1988).
Citation: Journal of Climate 22, 23; 10.1175/2009JCLI3145.1
