Variations in North Pacific sea surface temperature caused by Arctic stratospheric ozone anomalies

Recently, observations and simulations have shown that Arctic stratospheric ozone (ASO) variations affect the middle–high latitude tropospheric climate in the Northern Hemisphere. In particular, a connection from the ASO to El Niño–Southern Oscillation (ENSO) has been reported. However, no detailed study has been made of a key process in the connection, the influence of ASO on the North Pacific sea surface temperature (SST) and its underlying mechanism. Using observations, reanalysis and simulations, it is found that the ASO changes in March have the strongest connection with North Pacific SST variations in April. This implies a leading effect of ASO on North Pacific SST. The stratospheric circulation anomalies caused by March ASO changes can rapidly extend to the lower troposphere in the region 60°–90°N, 180°–120°W. Nevertheless, a theoretical analysis indicates that circulation anomalies from the region 60°–90°N, 180°–120°W in the lower troposphere would take about a month to propagate horizontally to the North Pacific middle latitudes (30°–60°N, 180°–120°W).


Introduction
The stratospheric circulation influences the chemical composition of the stratosphere, and anomalies in stratospheric circulation may also propagate downward to affect tropospheric weather and climate (Baldwin and Dunkerton 2001, Black 2002, Graf and Walter 2005, Scaife et al 2005, Sigmond et al 2008, Cagnazzo and Manzini 2009, Ineson and Scaife 2009, Gerber et al 2010, Reichler et al 2012, Ivy et al 2014, Karpechko et al 2014, Kidston et al 2015, Zhang et al 2016. Stratospheric ozone is vital to protecting life on the Earth, as it absorbs harmful solar ultraviolet radiation (Lubin andJensen 2002, Chipperfield et al 2015), and is also essential to the control of the stratospheric circulation via atmospheric radiative heating (Haigh 1994, Ramaswamy et al 1996, Forster and Shine 1997. A strong trend and a large variability in stratospheric ozone, which are associated with chemical and dynamical processes, respectively, mainly occur at high latitudes in both hemispheres. In the Southern Hemisphere, Antarctic stratospheric ozone has significantly decreased from the industrial revolution to the early 21st century as a consequence of anthropogenic emissions of ozone-depleting substances (Solomon 1990, 1999, Ravishankara et al 1994, 2009). Recent studies have found that the Antarctic ozone hole can influence the tropospheric climate in the Southern Hemisphere (Son et al 2008, 2010, Feldstein 2011, Kang et al 2011, Thompson et al 2011, Gerber and Son 2014, Waugh et al 2015; for example, rainfall, and sea surface temperature (SST).
Variability in Arctic stratospheric ozone (ASO) differs from that in Antarctic stratospheric ozone. On one hand, the Antarctic stratospheric ozone loss from 1950 to 2000 has been much larger than that of ASO (WMO 2011), as the winter/spring Antarctic polar vortex is much colder and stronger than the Arctic polar vortex. Accordingly, the effect of ASO loss on Northern Hemisphere surface climate may be less evident (e.g. Thompson and Solomon 2005). On the other hand, the amplitude of the year-to-year variability in ASO in the past several decades is at least as large as, or even larger than, that in the Antarctic (Randel 1988, Manney et al 2011, 2016, owing to frequent 'sudden warming' events in the high latitude Northern Hemisphere (Charlton and Polvani 2007). Given the noticeable difference in variability between ASO and Antarctic stratospheric ozone, their influences on the tropospheric climate may also be significantly different and merit further investigation. The impact of ASO interannual variations on Northern Hemisphere climate is one such example.
To investigate the possible surface impacts associated with extreme Northern Hemisphere ozone anomalies, Karpechko et al (2014) and Cheung et al (2014) used the ECHAM5 atmospheric circulation model and the UK Met Office operational weather forecasting system, respectively. They concluded that stratospheric ozone changes alone did not appear to have a significant effect on surface conditions on the synoptic scale. However, on the climatic scale, ASO variations caused anomalies of Northern Hemisphere mid-high latitude tropospheric circulation and sea level pressure (SLP). Smith and Polvani (2014) found significant influence of ASO on tropospheric circulation and the surface temperature and precipitation patterns using ensemble simulations. More interestingly, these impacts have very clear regional patterns. Subsequently, Calvo et al (2015) used a comprehensive stratosphere-resolving atmospheric model coupled with ocean, sea ice, and land components, and interactive stratospheric ozone chemistry to explore the impacts of large springtime ozone anomalies in the Arctic stratosphere on the troposphere and surface. They found that extremely small changes in ASO can produce large and robust anomalies in tropospheric wind, temperature, and precipitation in April and May over large parts of the Northern Hemisphere. More recently, Ivy et al (2017) presented observational evidence for linkages between extreme ASO anomalies in March and Northern Hemisphere tropospheric climate in spring (March-April), suggesting that March stratospheric ozone is a useful indicator of spring averaged (March-April) tropospheric climate in specific regions of the Northern Hemisphere.
Using statistical analysis and simulations, Xie et al (2016) established a possible connection from the ASO to the El Niño-Southern Oscillation (ENSO) by combining two steps: a high-latitude stratosphere-totroposphere pathway and an extratropical-to-tropical climate teleconnection. This implies that the ASO radiative anomalies influence the North Pacific Oscillation (NPO), and then the anomalous NPO and induced Victoria Mode anomalies (Bond et al 2003, Ding et al 2015 link to the North Pacific circulation that in turn influences ENSO. Garfinkel (2017) pointed out that the mechanism for ASO modulation of ENSO still deserves thorough analysis; for example, Xie et al (2016) reported a statistical relation between the ASO and North Pacific SST about one month later, which was also shown by Ivy et al (2017). Many questions arise here. In which month does the ASO have the strongest connection with North Pacific SST? How long does it take for the anomaly in the polar stratosphere to influence subtropical SSTs, and what is the pathway for this influence? The goal of the present study is to answer these questions. Note that the mechanism for downward propagation of the stratospheric signal to the North Pacific is not discussed in detail here, as it is still under investigation (e.g. Garfinkel et al 2013, Kidston et al 2015).

Data and method
The ozone data used in this study are obtained from the Black lines in figure S1 in the supplementary information available at stacks.iop.org/ERL/12/ 114023/mmedia show the ASO anomaly variations from 1979-2015 in each month over a region limited to approximately 60 • N-90 • N and 150-50 hPa, the region where the variability and depletion of ozone concentration is most pronounced in the Northern Hemisphere (Manney et al 2011). The monthly anomaly of ozone concentration (after removing the climatological mean   figure S1). The zonal mean SWOOSH dataset is a merged record of stratospheric ozone and water vapor measurements taken by a number of limb sounding and solar occultation satellites (SAGE-II/III, UARS HALOE, UARS MLS, and Aura MLS instruments). Its meridional resolution is 2.5 • and it has 31 pressure levels from 300 to 1 hPa. The zonal mean satellite-based GOZCARDS dataset  is produced from high quality data from past missions (e.g. SAGE, HALOE data) as well as ongoing missions (ACE-FTS and Aura MLS). Its meridional resolution is 10 • with 25 pressure levels from the surface up to 0.1 hPa. The ASO variations from MERRA2 are in good agreement with those from SWOOSH data and GOZCARDS data (figure S1), particularly in spring, which is the focus of this study. However, there are many missing values in the SWOOSH data and GOZCARDS data, so the following analysis uses MERRA2 ozone data.
The SST and SLP data were obtained from the UK Met Office Hadley Centre for Climate Prediction and Research SST (HadSST) and SLP (HadSLP) field datasets, respectively. Geopotential height, zonal wind, and temperature data were obtained from ERA-Interim and NCEP2.
A threshold of ±0.2 ppmv, which is equal to the standard deviation of the ASO series, is applied to the time series in figure S1 to identify ASO decrease and increase events. We use composite analysis to check the changes of circulation during ASO decrease and increase events in March during the period from 1979 to 2015. The selected events are listed in table 1.
We calculated the statistical significance of the correlation between two auto-correlated time series using the two-tailed Student's t-test and the effective number (N ef f ) of degrees of freedom (DOF) (Bretherton et al 1999). For this study, N ef f was determined by the following approximation : where N is the sample size, and XX and YY are the autocorrelations of the two sampled time series, X and Y, respectively, at time lag j. In March, sunlight reaches the north polar stratosphere. A decrease (increase) in ASO will lead to less (more) warming of the north polar vortex such that the ASO variations are positively correlated with north polar stratospheric temperature (figure 2(a)). The cooler (warmer) north polar stratosphere strengthens (weakens) the temperature gradient from the tropics to the North Pole. According to the thermal wind relationship, this situation results in a stronger (weaker) polar vortex such that the ASO variations are negatively correlated with north polar stratospheric circulation changes ( figure 2(b)). An obvious feature of figure 2(b) is that the north polar stratospheric circulation anomalies extend to the north polar troposphere, even to the surface. Another feature is a wave train signal in the troposphere from the North Pole to northern lower latitudes.     1(c)). Figure 2 implies the possibility of a signal caused by ASO changes propagating to the North Pacific. Figure 2 depicts the processes that the stratospheric circulation anomalies caused by ASO changes  . Thus, one might hypothesize that some of the downward influences related to ASO change may be due to changes of SFW. However such a hypothesis is not totally correct: it is found that the SFW is more likely to occur in May (not in March) during both the ASO decrease and increase events (figure S2 in supplementary information). It means that the time of SFW occurrence is not significantly affected by the ASO anomalies events.

Propagation of Arctic stratospheric circulation anomalies to the North Pacific
To study in more detail the horizontal propagation of circulation anomalies, the ray paths of waves at 850 hPa generated by the disturbed circulation over the region 60 • -90 • N and 180 • -120 • W in March and April are shown in figure 4. The wavenumbers along these rays are between 1 and 3. The wave ray paths represent the climate teleconnections; i.e. the propagation of stationary waves in realistic flows. The calculation of the wave ray paths and application of the barotropic model are described in detail by Li et al (2015) and Zhao et al (2015). We found that the Rossby waves generated by the disturbed circulation over the north polar lower troposphere mainly propagate southward to the central North Pacific after about a month in March and April (they propagate to the northern North Pacific in about 15 d). The wave ray paths are in good agreement with the composite analysis in figure 3. Figures 2, 3 and 4 imply that ASO changes take at least a month to influence North Pacific circulation and SST. Motivated by this finding, we recalculated the correlation coefficients between ASO changes in March and circulation anomalies over the North Pacific in April (figure 5). As expected, the correlation coefficients between March ASO and April zonal wind anomalies (figures 5(a) and (b)) are larger than between March ASO and March zonal wind anomalies (figures 2(b) and (c)). In addition, the correlation coefficients between March ASO and April SLP anomalies over the North Pacific are also more significant than those between March ASO and March SLP anomalies (figures 5(c) and 2(d)). The delayed effect of ASO on North Pacific circulation implies the delayed effect of ASO on North Pacific SST. Figure 6 is the same as figure 1, but with 15days (

Simulated lead of ASO on North Pacific SST
To further confirm the leading effect of ASO on North Pacific SST, we used the National Center for Atmospheric Research's Community Earth System Model (CESM) version 1.0.6 to simulate this process. CESM is a fully coupled global climate model that incorporates an interactive atmosphere (CAM/WACCM) component, ocean (POP2), land (CLM4), and sea ice (CICE). For the atmospheric component, we used the Whole Atmosphere Community Climate Model (WACCM), version 4 (Marsh et al 2013). WACCM4 is a climate model that has detailed middle-atmosphere chemistry and a finite volume dynamical core, and it extends from the surface to approximately 140 km. For our study, we disabled the interactive chemistry. WACCM4 has 66 vertical levels, with a vertical resolution of about 1 km in the tropical tropopause and lower stratosphere layers. Our simulations used a horizontal resolution of 1.9 • × 2.5 • (latitude × longitude) for the atmosphere, and approximately the same for the ocean.
The transient experiment (E 1 ) performed by CESM with the fully coupled ocean incorporating both   , and specified ozone forcing derived from the CMIP5 ensemble mean ozone output. E 1 is a historical simulation covering the period 1955-2005. All the forcing data used in this study are available from the CESM model input data repository.
The experiment E 1 covering the period 1955-2005 and with the specified ASO forcing applied to the CESM captures the leading effect of the specified ASO anomalies on the North Pacific ( figure 7). The circulation and VM-like pattern SST anomalies that appear over the North Pacific in April (figures 7(c) and (d)) are evidently larger than in March (figures 7(a) and (b)). This simulated result is similar to the observations. Note that the ozone forcing is specified in the simulation; therefore, the relationship between ASO and SST could only be caused by North Pacific SST anomalies related to the ASO changes.

Conclusions and discussion
The North Pacific SST anomalies are well known to significantly influence the Northern Hemisphere middle-high latitude climate. Thus, it is important to understand the factors that affect North Pacific SST. This study investigates the effect of ASO changes on North Pacific SST and the processes by which the stratospheric circulation anomalies caused by ASO changes propagate to the North Pacific, induce North Pacific Oscillation (NPO) anomalies, and then force North Pacific SST anomalies. Using observations, reanalysis data, statistical methods and simulations, it is found that the stratospheric circulation anomalies caused by March ASO changes rapidly extend all the way to the lower troposphere in the region 60 • N-90 • N, 180 • W-120 • W; however, the circulation anomalies from this region take about 30 days to propagate to the North Pacific middle latitudes (30N • -60 • N, 180 • W-120 • W). The horizontal transport of circulation anomalies is confirmed by wave ray tracing. This implies a leading effect of ASO on North Pacific SST. That is, the ASO changes in March have the strongest connection with North Pacific SST variations in April, not March.
This study still leaves a question unanswered. We found that the ASO changes lead North Pacific SST by about a month. It is known that North Pacific SST variations lead ENSO by about 12 months (Ding et al 2015). This cannot explain the 20 month lead of ASO changes on ENSO found by Xie et al (2016). One possible reason is that the North Pacific SST anomaly forced by the ASO is initially stored beneath the surface and is then re-entrained into the mixed layer when it deepens again in the following winter (Alexander et al 1999, Li 2012, Garfinkel 2017). The winter-to-winter persistence of North Pacific SST would influence the ENSO. This process takes about 20 months, and is the subject of our next study.