Paper 8 Imp

RESEARCH ARTICLE Association Between Extreme Atmospheric Anomalies Over
10.1029/2021JD036121
Antarctic Sea Ice, Southern Ocean Polar Cyclones and
Key Points:
• I ntense cyclones have a higher
Atmospheric Rivers
association with extreme temperature Ehlke Hepworth1 , Gabriele Messori2,3 , and Marcello Vichi1,4
anomalies than atmospheric rivers
(ARs) 1
Department of Oceanography, University of Cape Town, Cape Town, South Africa, 2Department of Earth Sciences, Centre
• Approximately half of the ARs are
of Natural Hazards and Disaster Science (CNDS), Uppsala University, Uppsala, Sweden, 3Department of Meteorology,
in the vicinity of extreme moisture
anomalies, while there is a weak Bolin Centre for Climate Research, Stockholm University, Stockholm, Sweden, 4Marine and Antarctic Research Centre for
link between cyclones and extreme Innovation and Sustainability, University of Cape Town, Cape Town, South Africa
moisture anomalies
• If an AR is in the vicinity of an
extreme temperature anomaly, there Abstract This study analyses the association of Southern Ocean extratropical cyclones and atmospheric
will likely be a concurrent extreme
moisture anomaly rivers (ARs) with extreme temperature and/or moisture atmospheric anomalies over Antarctic sea ice. The
hypothesis we test is whether the circulations associated with cyclones and ARs may routinely lead to the
presence of unusually warm, moist air masses over ice-covered regions. The analysis is conducted over the
Correspondence to:
extended Austral winter seasons (May-September) between May 1979 and September 2012, based on the
E. Hepworth,
[email protected] European Centre for Medium-Range Weather Forecasts Interim reanalysis data. Approximately 27% of intense
Southern Ocean cyclones and 20% of ARs occur in the vicinity of extreme temperature anomalies, while 12%
Citation: of intense cyclones and 46% of ARs occur in the vicinity of extreme moisture anomalies. We summarize our
Hepworth, E., Messori, G., & Vichi, M. results as follows: (a) extreme atmospheric anomalies over sea ice often occur in the absence of cyclones
(2022). Association between extreme or ARs; (b) intense cyclones have a stronger association with extreme temperature anomalies than ARs; (c)
atmospheric anomalies over Antarctic
approximately half of the ARs are in the vicinity of extreme moisture anomalies, while the latter's link with
sea ice, Southern Ocean polar cyclones
and atmospheric rivers. Journal of cyclones is weak; and (d) if an AR is in the vicinity of an extreme temperature anomaly, there will likely be a
Geophysical Research: Atmospheres, concurrent extreme moisture anomaly. This points to a strong association between ARs and moisture extremes,
127, e2021JD036121. https://doi.
and a nuanced link between Southern Ocean polar cyclones and atmospheric anomalies over Antarctic sea ice.
org/10.1029/2021JD036121
Received 4 NOV 2021

Accepted 6 MAR 2022 1. Introduction
The ocean-ice-atmosphere dynamics of the Antarctic and sub-Antarctic regions continue to test our understand-
ing both at synoptic and climate timescales. Specific scientific challenges include reducing biases in ocean-ice
models forced by atmospheric reanalyses (e.g., in ice motion; Massonnet et al., 2011; Lecomte et al., 2016; Uotila
et al., 2014), and accurately representing Antarctic features (such as sea ice) in climate models (e.g., Barthélemy
et al., 2018). Notwithstanding recent work in advancing our understanding of ocean-ice-atmosphere dynamics
(e.g., Bromwich et al., 2020), progress is hampered by the intrinsic difficulty of acquiring accurate measure-
ments in the high southern latitudes, especially in winter. Obstacles range from hostile environmental condi-
tions (Grieger et al., 2018; Smith et al., 2019), to extreme weather phenomena (e.g., Kwok et al., 2017; Matear
et al., 2015; Raphael, 2007), to the need to account (from an observational viewpoint) for the complex matrix
of synoptic interactions between ocean, sea ice, and atmosphere (e.g., Vichi et al., 2019). These interactions are
largely specific to the southern polar region.
In contrast with the Arctic Ocean, the Southern Ocean (SO) is host to frequent intense extratropical (polar)
cyclones, and strong baroclinicity (Simmonds et al., 2003). The SO cyclones are some of the most severe storms
on Earth and play a potentially large—but scarcely understood—role in modulating the ocean-ice-atmosphere
interplay (Godfred-spenning & Simmonds, 1996; Simmonds, 2015; Simmonds & Li, 2021; Vichi et al., 2019),
and, more broadly, the regional atmospheric circulation (Simmonds, 2003; Uotila et al., 2011). Over the satellite
period, the baroclinicity at the poles has increased (coupled with a decrease in baroclinicity at the midlatitudes)
resulting in a poleward shift of the cyclone tracks (Simmonds & Li, 2021). The total number of SO cyclones
calculated within a given area varies depending on the cyclone-tracking algorithm used (e.g., Grieger et al., 2018;
Messmer & Simmonds, 2021; Simmonds & Li, 2021). However, irrespective of the tracking algorithm, there is
robust agreement on the cyclones' climatological lifecycle. Cyclogenesis primarily takes place in the mid-latitude
© 2022. American Geophysical Union. Atlantic and Pacific Ocean sectors. The SO synoptic cyclones then typically follow a southeastward trajectory,
All Rights Reserved. and undergo cyclolysis closer to the Antarctic continent (Keable et al., 2002; Simmonds et al., 2003; Yuan
HEPWORTH ET AL. 1 of 15
21698996, 2022, 7, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2021JD036121 by Monash University, Wiley Online Library on [20/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Journal of Geophysical Research: Atmospheres 10.1029/2021JD036121
et al., 2009), with maximum cyclone density found in the higher-latitude Atlantic and Indian Ocean sectors
(Grieger et al., 2018; Hoskins & Hodges, 2005; Simmonds & Keay, 2000a, 2000b; Yuan et al., 2009). The
cyclones' radii commonly range between 500 and 2,000 km (Hoskins & Hodges, 2005; Uotila et al., 2011; Yuan
et al., 2009), with the more intense systems typically measuring approximately 4°–6° latitude (∼330–665 km;
Uotila et al., 2011). The radius of a cyclone does, however, depend on the phase of the cyclone's life cycle
(Simmonds, 2000), as well as its geographical location (Lim & Simmonds, 2007). There is large case-to-case
variability observed in the literature: for example, an explosive cyclone described by Vichi et al. (2019) during
the austral winter of 2017 displayed a radius in excess of 1,000 km.
A related, yet distinct atmospheric feature that can affect the ocean-ice-atmosphere dynamics are atmospheric
rivers (ARs). ARs are synoptic-scale long and narrow corridors of water vapor, often originating from tropical
moisture sources (Nieman et al., 2008). They account globally for ∼90% of the meridional water vapor trans-
port toward the mid and high-latitudes (Nash et al., 2018; Wille et al., 2021), acting as key contributors to the
interannual variability of precipitation across sea ice (Wille et al., 2021). In the Southern Hemisphere, ARs are
predominantly located between 40° and 60°S and can extend poleward achieving lengths in excess of 2,000 km
(Simmonds et al., 2012). ARs are typically identified ahead of the cyclone's cold front (Francis et al., 2020; Nash
et al., 2018; Zhu & Newell, 1998), which is a region of enhanced water vapor content (Nieman et al., 2008).
Intense ARs can in turn aid in the strengthening of cyclones by providing more water vapor for latent heat release
(Zhang et al., 2019). Although it is common for ARs to be associated with cyclones, it is relatively not as common
for cyclones to be paired with an AR (Zhang et al., 2019).
A key role of both SO cyclones and ARs is their contribution to the poleward energy transport (Patoux et al., 2009),
which can be associated with large moisture and temperature anomalies in the high latitudes (Francis et al., 2020;
Grieger et al., 2018). For example, Krinner et al. (2007) presented observations indicating that frequent, strong
cyclones off the Antarctic coast bring the bulk of the annual total precipitation. A high match between cyclones
and precipitation is shown by Papritz et al. (2014), demonstrating that, over the SO, between 60% and 90% of
strong precipitation (>75th percentile) events are a result of cyclones and fronts.
However, an analysis of the concurrent relationship of extratropical cyclones and ARs with ice-covered Antarc-
tic regions has not been attempted before in the literature. Warm-core anomalies have been identified in five
cyclones crossing the winter South Atlantic marginal ice zone (MIZ), supported by in-situ measurements on sea
ice (Vichi et al., 2019). ARs have further proven to play a key role for understanding the Antarctic surface mass
balance (e.g., Francis et al., 2020; Gorodetskaya et al., 2014). For example, the anomalously high amounts of total
precipitable water and cloud liquid water content (as a result of ARs) were a key contribution to the sea ice melt
in the observed Weddell Polynyas in November 1973 and September 2017 (Francis et al., 2020). These findings
are in agreement with Wille et al. (2019) who indicated that surface melt events in the West Antarctic are linked
with ARs and the related anomalous poleward moisture transport. These results support the hypothesis that it is
common for ARs and intense cyclones to engender large temperature and/or moisture anomalies over the south-
ern high latitudes, and that these in turn can affect sea ice.
Previous studies have also highlighted the role of other dynamical processes in modulating extreme atmospheric
anomalies over the SO. These include the El-Nino-Southern Oscillation and the related Southern Annular Mode
(Hobbs et al., 2016; Stueker et al., 2017), atmospheric Rossby waves propagating to the high-latitude southeast
Pacific (Ding & Steig, 2013; Turner, 2004), and the semiannual oscillation and Antarctic circumpolar wave
(e.g., Simmonds, 2003). Similar to the synoptic-scale features, analyses specifically dedicated to sea-ice covered
regions are largely lacking. As a first step to address this broad knowledge gap, and to keep our analysis focused,
we test the hypothesis that the circulations associated with cyclones and ARs may routinely lead to the presence
of unusually warm, moist air masses over sea-ice covered regions of the SO. Specifically, we track cyclones and
ARs located over or in the vicinity of sea ice throughout their life-cycle, and analyze extreme atmospheric anom-
alies over the sea ice and their association with the presence of cyclones and ARs.
In Section 2, the data and methods used to assess the link between the two synoptic systems (cyclones and ARs) and
extreme atmospheric anomalies are outlined. We consider sensitivity to the distance between the synoptic systems
and extreme atmospheric anomalies, and to the intensity of cyclones. This enables a comprehensive characterization
of the relationship that both cyclones and ARs have with extreme temperature and moisture anomalies over Antarc-
tic sea ice (Section 3), which is not yet quantified in the literature. Concluding remarks are presented in Section 4.
2. Data and Methods

2.1. Data
We based our analysis on the European Centre for Medium-Range Weather Forecasts (ECMWFs) Interim
Re-Analysis (ERA-Interim; Dee et al., 2011) over all extended Austral winter seasons (May-September; MJJAS)
between May 1979 and September 2012. ERA-Interim provides 6-hourly data at a horizontal resolution of 0.75°.
ERA-Interim was used because the cyclone tracking method used for our analysis (see Section 2.3) was developed
using the ERA-Interim mean sea level pressure (MSLP). Consistency was thus provided throughout this study by
using the same database for the sea ice and atmospheric variables (Section 2.2). In addition, recent work has shown
that ERA-Interim and its successor, ERA5, have similar skill in reproducing the observed atmospheric conditions
associated with intense cyclones over the Antarctic MIZ (Vichi et al., 2019). Similarly, McDonald and Cairns (2020)
compared the consistency in surface winds in eight reanalysis products and found that very similar patterns resulted
from the ERA-Interim output and other reanalyses (including ERA5). McDonald and Cairns (2020) continued to
recommend using either ERA-Interim or ERA5 over their region of interest, the Ross Sea.
2.2. Atmospheric Anomalies
Our analysis of extreme atmospheric temperature and moisture anomalies focuses on the Antarctic sea ice region,
defined here as the region where the sea ice concentration (SIC) is above 15% (Meier & Stroeve, 2008). Temper-
ature 2 m above sea level (t2m) and total column water were used to locate positive temperature and moisture
anomalies. These atmospheric anomalies were computed as deviations from their respective climatologies, which
were calculated as grid-cell averages over 1979–2012, with a 6-hourly temporal resolution.
From these results, the 95th percentile of the anomalies was calculated and the values above this threshold were
defined as extreme atmospheric anomalies (which we, hereafter, refer to as “extremes”). Figure 1 shows an exam-
ple of temperature and moisture extremes identified over sea ice for 30 July 1990 at 06h00 UTC.
Atmospheric extremes are typically detected in clusters of contiguous grid cells (Figure 1c). Each cluster is iden-
tified, and the longitude and latitude of its centroid are used to define the location of the extremes (Figure 1d).
If the distance between two or more extreme anomaly centroids is below a given threshold, the extremes are
grouped together, and their location is defined as the average of the centroids. We tested threshold distances of
400, 600, and 1,000 km, depending on the size of the search radius between a cyclone and an atmospheric anom-
aly (as explained in Section 2.3).
Separate 95% thresholds to define the atmospheric extremes were initially tested for the outer and inner sea-ice
regions (which were defined as the regions of relatively higher or lower sea-ice variability, respectively). However,
we observed that separating these regions provided a discontinuous distribution (not shown) and, thus, a single
threshold for the entire Antarctic sea ice region was used to produce a more spatially coherent distribution of
atmospheric extremes.
2.3. Detection of Cyclone and Atmospheric River Tracks
The present study uses the tracking algorithm developed by Pinto et al. (2005) which identifies cyclone core
tracks using vorticity calculated from the Laplacian of the MSLP. The tracks have been computed from ERA-In-
terim reanalysis at 6-hourly temporal resolution, as provided in the intercomparison by Neu et al. (2013) and
analyzed by Grieger et al. (2018). Since consistency across different cyclone-tracking methods is generally higher
for intense cyclones than for shallow ones (Neu et al., 2013), we expect our analysis not to be overly sensitive to
the choice of the tracking algorithm. We discuss this aspect further in Section 4.
To identify ARs, the Image-Processing based Atmospheric River Tracking (IPART) algorithm, made available
by Xu et al. (2020), was used. This algorithm was used to detect and track ARs from gridded Integrated Vapor
Transport data from ERA-Interim reanalysis at 6-hourly temporal resolution. In contrast with cyclones, ARs are
not identified by an unambiguous core. Since we are interested in anomalies over sea ice, the southernmost point
of each AR was used as its location. The intensity of ARs was not considered in this analysis.
Figure 1. Atmospheric (a) temperature and (b) moisture anomalies (over Antarctic sea ice), from the ERA-Interim data set
on 30 July 1990 at 06h00 UTC, from which (c) temperature extremes (red contours) and moisture extremes (green contours)
were extracted, and (d) the location of each anomaly's centroid was calculated. The sea-ice edge is denoted by the black
contour lines.
2.4. Search-Radius Method and Random Association
The location of each atmospheric extreme's centroid was extracted (see Section 2.2), and the number of cyclones
and ARs within a given radius of the centroid was calculated (which we term search-radius method). We then
considered: (i) the distance that a cyclone or AR was from an atmospheric extreme; and (ii) the definition of
cyclone intensity.
1. W hen considering distance, the search-radius method was repeated using a radius (considered as the impact-
ing area of individual cyclones or ARs) of 400, 600, or 1,000 km. Therefore, the cyclones and ARs included
in this analysis were those within a given radius of an atmospheric extreme centroid on sea ice. The percentage
of cyclones or ARs within a given radius from an extreme, relative to the total number of cyclones or ARs
tracked, was calculated. The total number of cyclones or ARs tracked was defined as their respective total
numbers over the sea ice or north of the sea ice but within one search radius from the sea ice edge.
2. When considering cyclone intensity, the search-radius method was applied separately for intense cyclones
as well as non-intense cyclones. There are numerous ways to define intense cyclones. For example, consid-
ering the MSLP climatological field, and defining intensity in terms of “relative” central pressure (e.g., Lim
& Simmonds, 2002; Simmonds & Wu, 1993), or in terms of percentile thresholds (Grieger et al., 2014).
This study classified intense cyclones as tracked cyclones (see point (i) above) with minimum central pres-
sure threshold below the 1st, 5th, or 10th percentiles of the distribution of all tracked cyclones over MJJAS.
Conversely, non-intense cyclones were identified as those with minimum pressure levels greater than the
chosen threshold. These thresholds were computed separately for each month (from May to September), with
an average minimum central pressure threshold of around 941, 952, and 957 hPa below the 1st, 5th, and 10th
percentiles, respectively.
The results for these different thresholds were compared. In Section 3.1, an atmospheric extreme is considered to
be associated with a cyclone or with an AR if its centroid is within a search radius (400, 600, or 1,000 km) from the
cyclone or AR's location. In Section 3.2, we focus on the results using the 600 km radius and the 5th percentile thresh-
old. This is because very large cyclones (with a radius larger than approximately 6° latitude [∼665 km]) are typically
not very intense (Uotila et al., 2011), and the 5th percentile provides a good balance between extremeness of the
cyclones and sample size. We nonetheless show the results for the two additional search radii (400 and 1,000 km) and
intensity thresholds (1st and 10th percentiles) to verify the robustness of our conclusions to the chosen parameters.
Additionally, to validate the strength of the relationship between the extremes and synoptic systems, we imple-
mented a random sampling procedure. This was done by randomly drawing the same number of grid cells as the
number of tracked ARs or intense cyclones (within the 5th percentile threshold). This was repeated 2,000 times,
and the 95th percentile of the number of grid cells that were within 600 km from the extremes in each one of the
2,000 random draws was calculated. These results provide an indicative upper confidence interval for a statisti-
cally significant link between the synoptic features and on-ice extremes.
3. Results
3.1. Multifarious Links Between Cyclones, ARs, and Extreme Atmospheric Anomalies
To make an initial evaluation of the link between SO cyclones, ARs, and extreme atmospheric temperature and
moisture anomalies, some illustrative case studies are presented.
Case study 1 (30 July 1990 at 06h00 UTC, Figure 2) shows that atmospheric extremes can be induced by intense
as well as non-intense cyclones, and that ARs are found on the leading side of a cyclone. The chosen timestep
Figure 2. Atmospheric (a) temperature, and (b) moisture anomalies shown over Antarctic sea ice (intermediate gray contour
lines) on 30 July 1990 at 06h00 UTC (case study 1). The extreme atmospheric (a) temperature and (b) moisture anomalies are
denoted with thick black contour lines. The MSLP isolines are overlain (thin black contour lines), and the ARs are denoted by
the elongated pink and purple features. All data are from the ERA-Interim data set.
Figure 3. As Figure 2, but for (a, b) 13 July 1999 at 18h00 UCT (case study 2), and (c, d) 6 August 1980 at 06h00 UTC (case
study 3).
displays both an intense cyclone (labeled “Cyclone 1” in Figure 2b) with a minimum pressure of 933 hPa and a
non-intense cyclone (labeled “Cyclone 2”) with a minimum pressure of 972 hPa. There are temperature and mois-
ture extremes on the eastward and poleward flank of Cyclone 1 in the East Atlantic Ocean as well as in the Ross
Sea in the vicinity of Cyclone 2. Additionally, on the leading side of Cyclone 1, there is an AR extending pole-
ward, while no AR is observed in Cyclone 2. After a visual inspection of several cases reveals that the extremes
are often located on the south-eastern flank, downstream of the cyclone's direction, with the temperature anomaly
extending over the core of the cyclones.
Case studies 2 (13 July 1999 at 18h00 UTC, Figures 3a and 3b) and 3 (6 August 1980 at 06h00 UTC, Figures 3c
and 3d) suggest that temperature and moisture extremes need not occur concurrently, that intense cyclones or
ARs do not necessarily engender on-ice extremes, and that such extremes can occur in the absence of cyclones
or ARs. In case study 2, in the Weddell and Bellingshausen sectors, two cyclones (labeled “Cyclone 1” and
“Cyclone 2” in Figure 3b) are associated with temperature extremes, while the co-located moisture anomalies
are positive but not extreme. Conversely, in case study 3, two cyclones are associated with moisture extremes in
the eastern Atlantic and Weddell sectors (labeled “Cyclone 3” and “Cyclone 4” in Figure 3d) but the temperature
anomalies in these sectors are not extreme. There is an AR on the leading side of an intense cyclone (in the 5th
Figure 4. As Figure 2, but for (a, b) 4 August 1994 at 06h00 UTC (case study 4) and (c, d) 18 September 2012 at 06h00
UCT (case study 5).
percentile range) in the East Atlantic sector, but neither the cyclone nor the AR are associated with extreme on-ice
anomalies (case study 2; Figures 3a and 3b). There are also ARs on the leading edges of non-intense cyclones
in the Indian Ocean (case study 2) and East Atlantic (case study 3) sectors (Figures 3a–3d, respectively). In both
cases, the non-intense cyclones and ARs are associated with small-scale moisture extremes but not temperature
extremes. Although Cyclones 1 and 4 are associated with a temperature and moisture extreme, respectively, there
is no AR present on the leading side of these cyclones (Figures 3a and 3d, respectively). Finally, case study 3 also
shows that one may have extensive regions of temperature extremes without any cyclone or AR in the vicinity
(Figure 3c).
Case studies 4 (4 August 1994 at 06h00 UTC, Figures 4a and 4b) and 5 (18 September 2012 at 06h00 UTC,
Figures 4c and 4d) highlight the fact that temperature and moisture extremes can be induced by cyclones both
close to the ice edge as well as deep into the ice. For example, moisture extremes are associated with Cyclones 2
and 3 (labeled in Figures 4b and 4d, respectively) which are close to the ice edge (155 and 120 km from the edge,
respectively). On the opposite, a temperature extreme is associated with Cyclone 1 (labeled in Figure 4a) which is
deep into the ice (1,095 km from the ice edge), and concurrent temperature and moisture extremes are associated
with Cyclone 4 (1,200 km deep into the sea ice; labeled in Figure 4d). Furthermore, in relation to Cyclone 4,
which affects the extended Ross Sea ice surface, the warm-core extreme penetrates deep into the ice, while the
Figure 5. As Figure 2, but for (a, b) 10 July 1982 at 06h00 UTC (case study 6), and (c, d) 11 July 2009 at 00h00 UTC (case
study 7).
moisture extreme is confined to the ice edge. A contrasting location of the temperature and moisture extremes
relative to the cyclone's core is thus observed. There are ARs on the leading sides of Cyclones 2 and 4 which in
both cases may be visually associated with the presence of moisture extremes (Figures 4b and 4d).
To conclude our qualitative review, case studies 6 (10 July 1982 at 06h00 UTC, Figures 5a and 5b) and 7 (11
July 2009 at 00h00 UTC, Figures 5c and 5d) expand on case studies 2 and 3 in showing that extremes are not
always found in correspondence with cyclones or ARs, and that even cyclones belonging to the high intensity
categories can engender little or no atmospheric extremes. For example, in case study 6, a temperature extreme
in the Amundsen/Bellingshausen sector is observed with no cyclone or AR in the vicinity, and an intense cyclone
(in the 5th percentile range; labeled “Cyclone 1” in Figure 5b), with a relatively small AR on its leading side in
the East Atlantic Ocean, engenders no extremes. Similarly, in case study 7, there is an intense cyclone (in the
1st percentile range; labeled “Cyclone 2” in Figure 5d) in the Indian Ocean sector, with an AR, leading to only
limited atmospheric extremes, while the more extensive extremes in the East Pacific sector are not located in the
vicinity of a cyclone nor of an AR.
Figure 6. The total number of (a) non-intense cyclones, (b) intense cyclones, and (c) atmospheric rivers over Antarctic sea ice and 600 km north of the ice edge, and
(d) temperature and (e) moisture extremes over Antarctic sea ice from 1979 to 2012 over the extended Austral winter period of May-September. The cyclone data are
from the cyclone-tracking method developed by Pinto et al. (2005), the AR data are from the IPART algorithm, made available by Xu et al. (2020), and the extreme
atmospheric data is from the ERA-Interim data set. The total number of cyclones and ARs is computed per grid cell on the ERA-Interim grid and smoothed using a
Gaussian Filter (with σ = 2).
3.2. Distribution of Synoptic Features and Extreme Atmospheric Anomalies
The case studies presented above indicate that the association that cyclones and ARs have with atmospheric
extremes is complex. To gain a more systematic understanding, we next analyze the full data set used in this study.
As documented by for example, Simmonds and Keay (2000a, 2000b), Hoskins and Hodges (2005), Yuan
et al. (2009), and Grieger et al. (2018), the highest numbers of SO cyclones are found in the Atlantic and Indian
Ocean sectors. Non-intense cyclones and ARs, which are largely grouped in the Atlantic and eastern Antarctic
sectors, follow this geographical distribution. Moreover, there is a hotspot of both non-intense cyclones and ARs
in the eastern Pacific sector (Figures 6a and 6c, respectively). The intense cyclones instead cluster in three hotspot
areas, namely the eastern Atlantic sector, western Pacific sector, and the Ross/Amundsen sector (Figure 6b).
The atmospheric extremes, detected as described in Section 2.2, present yet a different distribution. Overall,
the extremes are mostly located in the western Antarctic region with the Ross/Amundsen sector emerging as
a hotspot, especially for temperature extremes (Figures 6d and 6e). Temperature extremes and cyclones are
predominantly found closer to the Antarctic continent (Figures 6a, 6b and 6d) while moisture extremes and ARs
are predominantly located further away from the continent and closer to the sea-ice edge (Figures 6c and 6e).
To gain a more in-depth understanding of the difference between intense and non-intense cyclone distributions,
we now turn our attention to the identification of cyclones found within a given search radius from the extremes'
Figure 7. The total number of Southern Ocean (a-c) non-intense and (d-f) intense cyclones within a 600 km radius of extreme atmospheric anomalies: (a, d)
temperature, (b, e) moisture, and (c, f) concurring temperature and moisture. The total number of cyclones is computed per grid cell on the ERA-Interim grid and
smoothed using a Gaussian Filter (with σ = 2).
centroids (see Section 2.4). Temperature extremes are frequently associated with both intense and non-intense
cyclones close to the Antarctic continent (Figures 7a and 7d), whereas moisture extremes are primarily associated
with those close to the sea-ice edge (Figures 7b and 7e). Temperature extremes appear less sensitive to the inten-
sity of cyclones than moisture extremes (see also Figure 8), and thus are found in the regions where cyclones are
modally located: the high number of non-intense cyclones in the eastern Antarctic sector (Figure 6a) and the three
hotspot regions of the intense cyclones (Figure 6b) emerge clearly in Figures 7a and 7d, respectively. Moisture
extremes show a frequent association with non-intense cyclones in the eastern Pacific sector (Figure 7b), whereas
the moisture extremes associated with intense cyclones are observed chiefly in the mid-Atlantic and Indian Ocean
sectors (specifically located around 0° and 90°E, respectively; Figure 7e).
To quantify the frequency with which (non-intense or intense) cyclones and ARs are associated with temper-
ature and moisture extremes, we computed the percentage of cyclones and ARs within selected radii from the
atmospheric extremes (Figure 8). The histogram bars highlight the results specific to the 600 km search radius
and the non-intense cyclones (greater than the 5th percentile threshold), the intense cyclones (within the 5th
percentile range), and the ARs. Additional search radii and percentile ranges are shown by the thin range bars
Figure 8. Percentage of the non-intense cyclones (yellow bars), intense cyclones (green bars), and ARs (purple bars) within
600 km of extreme atmospheric temperature anomalies, moisture anomalies, and concurrent temperature and moisture
anomalies. The results for the extreme cyclones defined using the 1st and 10th percentile thresholds are denoted by the
upper and lower limits of each range bar, respectively, for the 400 km (dark blue range bars), 600 km (black range bars), and
1,000 km (light blue range bars) search radii. The AR 1,000 and 400 km search radii are denoted by the upper and lower pink
range bars. The horizonal bars denote an indicative upper confidence interval for our random sampling.
and illustrate the sensitivity of the relationship that cyclones and ARs have with atmospheric extremes to the
chosen parameters.
The results for the 600 km radius (Figure 8, Section 2.4) show that approximately 16% of the non-intense SO
cyclones, 27% of the intense cyclones, and 20% of the ARs are in the vicinity of temperature extremes. Similar
percentages of both intense and non-intense cyclones (12%–13%) are within 600 km of moisture extremes, while
46% of ARs are within 600 km of moisture extremes. Only 6% of either non-intense or intense cyclones, and 19%
of ARs are in the vicinity of both a temperature and moisture extreme. The highest percentage of cyclone cores
in the vicinity of atmospheric extremes expectedly occurs when the search radius is the greatest (1,000 km) and
the intense cyclones are below the lowest percentile threshold (1st percentile). For this combination of param-
eters, approximately 57% of intense cyclones can be associated with temperature extremes, 35% with moisture
extremes, and 26% with concurrent temperature and moisture extremes.
The 95th percentile of our random sampling (see Section 2.4) suggests that there is approximately a 12%–13%
chance that an intense cyclone or an AR will be 600 km from an extreme temperature or moisture anomaly. This
decreases to approximately a 5% for concurrent temperature and moisture extremes.
4. Discussion and Conclusions

Vichi et al. (2019) presented an observational case study where an explosive cyclone crossing the East Antarctic
MIZ in July 2017 engendered large atmospheric temperature anomalies over the sea ice that propagated toward
the Antarctic continent. This observation resonates with previous work that highlighted the role of SO cyclones
in contributing to the average poleward transport of energy and moisture (Patoux et al., 2009), and in driving
large temperature and moisture anomalies (e.g., Krinner et al., 2007; Lynch et al., 2006; Papritz et al., 2014;
Uotila et al., 2011). However, a systematic classification of this relationship has not been attempted before.
Additionally, it is known that ARs (which are typically found on the leading side of cyclones) play a large role
in transporting heat and moisture poleward over sea ice (Francis et al., 2020). Recent studies have shown the
regional-scale impact that ARs and large moisture influxes have within the SO and in the Antarctic region (e.g.,
Gorodetskaya et al., 2014; Francis et al., 2020; Wille et al., 2019). We therefore adopted the working hypothesis
that it is common for ARs and intense cyclones to engender extreme temperature and/or moisture anomalies over
the ice-covered regions of the SO, and have tested this using reanalysis data and algorithms which track cyclones
and ARs. In our methodology, an extreme atmospheric anomaly was associated with a cyclone or an AR if it was
within a 600 km radius of the synoptic system's location. The percentage of association changed with the choice
of the search radius, but the conclusions are robust to reasonable alterations of this threshold. We evaluate the
strength or weakness of the associations relative to a random sampling procedure which indicates that, given our
methodology and parameter choices, one may take a 12%–13% chance of an AR or intense cyclone being in the
vicinity of an extreme as a reasonable confidence bound.
The association of intense cyclones or ARs with temperature extremes over sea ice are, respectively, relatively low
and moderate, and there is a weak link between cyclones and moisture extremes. This is seen in Figure 8 where
approximately one sixth (16%) of non-intense cyclones, a quarter (27%) of intense SO cyclones, and a fifth (20%)
of ARs are in the vicinity of extreme temperature anomalies. This is in line with results presented by Sinclair and
Dacre (2019) who showed that moisture transport to Antarctica was less dependent on cyclone intensity and more
dependent on the translation speed of the cyclone. We also note that when considering the most intense cyclones
and an expanded search radius around them, over half of them (57%) match a temperature extreme, and about
one third (35%) match a moisture extreme, and about a quarter (26%) match concurrent temperature and moisture
extremes. Additionally, our results show a strong link between ARs and moisture extremes, as approximately
half of the ARs (47%) are in the vicinity of a moisture extreme. It is interesting to observe that 20% of ARs are
600 km from temperature extremes, and 19% are 600 km from concurrent temperature and moisture extremes.
This indicates that if an AR is in the vicinity of a temperature extreme, there will likely be a concurrent moisture
extreme. The strong link between ARs and moisture extremes is additionally highlighted when considering the
1,000 km search radius: 26% of ARs are 1,000 km from temperature extremes and 28% are 1,000 km from
concurrent temperature and moisture extremes. This slight increase suggests that, when increasing the search
radius to 1,000 km, it is possible for an AR to concurrently be in the vicinity of a temperature extreme and more
than one moisture extreme. Our main conclusions can be summarized as follows: (a) extreme atmospheric anom-
alies over sea ice often occur in the absence of cyclones and ARs; (b) intense cyclones have a stronger association
with temperature extremes than ARs; (c) approximately half of the ARs are in the vicinity of moisture extremes,
while the latter's link with cyclones is weak; and (d) if an AR is in the vicinity of a temperature extreme, there will
likely be a concurrent moisture extreme. In other words, we find a moderate but significant association between
intense cyclones and temperature extremes, a strong association between ARs and moisture extremes, little to no
association between cyclones and moisture extremes, and that, relative to ARs, intense cyclones have a higher
association with temperature extremes.
Our working hypothesis was that it is common for intense cyclones and ARs to engender temperature and/or mois-
ture extremes over the SO, including over Antarctic sea ice. There indeed is a strong association between ARs and
moisture extremes over sea ice. However, although cyclones are considered to be a primary transport mechanism
of heat and moisture into the polar regions (particularly in winter; Grieger et al., 2018), the link between intense
cyclones and temperature and moisture extremes over Antarctic sea ice proves to be more complex than origi-
nally hypothesized. Our results do show that, relative to non-intense cyclones, intense cyclones are more likely
to induce extreme atmospheric anomalies (Figure 8). Nonetheless, the case study by Vichi et al. (2019) is only
representative of 57% of intense cyclones below the 1st percentile threshold, which is the most restrictive defi-
nition we considered here. Our results further highlight important differences between temperature and moisture
extremes. The geographical composites presented in Figure 7 show that temperature extremes are less dependent
on cyclone intensity than moisture extremes, as the former are found in regions where non-intense and intense
cyclones are modally located (Figures 7a and 7d, respectively). The moisture extremes associated with intense
cyclones are predominantly located in the mid-Atlantic and Indian Ocean sectors (Figure 7e), in agreement with
Grieger et al. (2018). Additionally, in the Ross/Amundsen sector, there is a relatively high number of intense
cyclones as well as extreme atmospheric anomalies (Figures 6b–6d). Our methodology cannot determine whether
atmospheric extremes in this region are due to the relatively higher number of intense cyclones, or, vice versa,
whether the extremes provide favorable conditions for intense cyclones surviving over the sea ice. Based on the
literature, the former may seem more plausible (e.g., Krinner et al., 2007; Lynch et al., 2006; Papritz et al., 2014).
Our results highlight the need to focus future work on regional scales, to verify whether, and through which phys-
ical mechanisms, intense cyclones are more or less likely to engender atmospheric extremes in specific sectors of
the SO. This includes clarifying whether the explosive nature of cyclones or details of their tracks and translation
speed may affect their ability to generate on-ice atmospheric extremes.
We can also leverage our analysis to make hypotheses about the role of sea ice in modulating these extremes.
We find that extreme temperature anomalies over sea ice are mostly located closer to the Antarctic continent
(where sea ice is more consolidated with lower variability), and moisture extremes are largely observed closer to
the sea-ice edge (where sea-ice variability is higher; Figures 6 and 7d,e). We therefore speculate that there is an
interactive role between the atmospheric extremes and sea-ice variability. Krinner et al. (2007) provides specific
regional-scale examples of how changes in SIC influence the number of cyclones which, in turn, leads to changes
in precipitation. Additionally, Francis et al. (2020) and Wille et al. (2019) indicate the impact ARs (bringing in
moisture) can have on the sea ice region. Our results support the hypothesis that extreme atmospheric anomalies
are, in turn, likely to contribute to sea ice conditions. Although this is an aspect which we did not address explic-
itly here, we identify it is a key area for future research activity. It is interesting to note that, although temperature
increases tend to lead to increased moisture transport due to the higher moisture holding capacity of warmer air,
an extreme temperature anomaly rarely occurs within the same cyclone as an extreme moisture anomaly. Addi-
tionally, only 19% of ARs are in the vicinity of both a temperature and a moisture extreme. It is, therefore, possi-
ble that many of the analyzed cases may lead to intense solid precipitation without extreme warming. This would
preserve the snow on the sea ice and extend the persistence of underlying ice. These factors may contribute to the
large sea-ice variability observed closer to the Antarctic sea-ice edge. Therefore, a stimulating analysis would be
to repeat the present study over regions of high and low sea-ice variability.
It is important to note that our analysis needs to be repeated using other atmospheric data sets and cyclone tracking
methods in order to investigate its robustness. For example, results presented by Grieger et al. (2018) and Messmer
and Simmonds (2021) show that the total number of SO cyclones vary depending on the cyclone tracking algo-
rithm used. The cyclone tracking algorithm by Pinto et al. (2005) that was used here, identifies a larger number of
tracks in the SO than other algorithms (Neu et al., 2013). While we do not expect this to have a large impact for
intense cyclones—where different algorithms generally show a good level of agreement (Grieger et al., 2018; Neu
et al., 2013)—it is likely to lower the percent association between non-intense cyclones and extreme atmospheric
anomalies. Indeed, given a fixed number of atmospheric extremes, an algorithm identifying more cyclone tracks
may intuitively lead to fewer of these being associated with the extremes. At the same time, the Pinto et al. (2005)
algorithm identifies roughly 20% more cyclones in the SO than the average of all algorithms considered in Neu
et al. (2013). As such, we deem that our qualitative conclusions concerning the fact that only a modest fraction of
non-intense cyclones engender extreme atmospheric anomalies would hold even if using algorithms identifying a
lower number of tracks. Additionally, it will be fruitful to repeat this analysis using a relative central pressure method
to identify intense cyclones (rather than the percentile threshold method we adopted). This is motivated by previous
studies that have shown how the dynamic meaning of the cyclone's central pressure can be better understood relative
to the background climatological MSLP (Lim & Simmonds, 2002). For example, Simmonds and Wu (1993) have
shown that cyclones become weaker (relative to the climatological MSLP) over most of the Antarctic sea ice region.
As alluded to in Section 1, previous studies have assisted in improving our understanding of ocean-ice-atmos-
phere dynamics in the Antarctic and subantarctic regions (e.g., Bromwich et al., 2020). Our results contribute to
this effort by providing a quantitative evaluation of the link between SO cyclones, ARs, and extreme temperature
and moisture over sea ice.
Data Availability Statement

ERA-Interim reanalyses were obtained from the ECMWF website, and the data for the cyclone tracks were from
the tracking algorithm developed by Pinto et al. (2005) (Method 02) and accessed at https://naturwissenschaften.
ch/organisations/proclim/activities/project_imilast/data_download. The IPART algorithm for the atmospheric
rivers was made available by Xu et al. (2020), and accessed at GitHub-ihesp/IPART:Image-ProcessbasedAtmos-
Acknowledgments
phericRiverTracking(IPART)algorithms.
This work is based on the research
supported by the National Research
Foundation of South Africa (Grant Nos. References
118598 and 112632) and the Swedish
Foundation for International Cooperation Barthélemy, A., Goosse, H., Fichefet, T., & Lecomte, O. (2018). On the sensitivity of Antarctic sea ice model biases to atmospheric forcing
in Research and Higher Education (Grant uncertainties. Climate Dynamics, 51(4), 1585–1603. https://doi.org/10.1007/s00382-017-3972-7
No. SA2017-7063). Computations were Bromwich, D. H., Werner, K., Casati, B., Powers, J. D., Gorodetskaya, I. V., Massonnet, F., et al. (2020). The year of polar prediction in the southern
performed using facilities provided by the hemisphere (YOPP-SH). Bulletin of the American Meteorological Society, 101(10), E1653–E1676. https://doi.org/10.1175/BAMS-D-19-0255.1
University of Cape Town's ICTS High Dee, D. P., Uppala, S. M., Simmons, A. J., Berrisford, P., Poli, P., Kobayashi, S., et al. (2011). The ERA-Interim reanalysis: Configuration
Performance Computing team: and performance of the data assimilation system. Quarterly Journal of the Royal Meteorological Society, 137656, 553–597. https://doi.
hpc.uct.ac.za. Lastly, the authors would org/10.1002/qj.828
like to acknowledge Diana Francis for Ding, Q., & Steig, E. J. (2013). Temperature change on the Antarctic Peninsula linked to the tropical Pacific. Journal of Climate, 26(19),
recommending the addition of ARs to 7570–7585. https://doi.org/10.1175/JCLI-D-12-00729.1
this study.
Francis, D., Mattingly, K. S., Temimi, M., Massom, R., & Heil, P. (2020). On the crucial role of atmospheric rivers in the two major Weddell
Polynya events in 1973 and 2017 in Antarctica. Science Advances, 6(46), eabc2695. https://doi.org/10.1126/sciadv.abc2695
Godfred-spenning, C. R., & Simmonds, I. A. N. (1996). An analysis of Antarctic sea-ice and extratropical cyclone associations. International
Journal of Climatology, 16, 1315–1332. https://doi.org/10.1002/(sici)1097-0088(199612)16:12<1315::aid-joc92>3.0.co;2-m
Gorodetskaya, I. V., Tsukernik, M., Claes, K., Ralph, M. F., Neff, W. D. M., & Van Lipzif, N. P. M. (2014). The role of atmospheric rivers in
anomalous snow accumulation in East Antarctica. Geophysical Research Letters, 41(17), 6199–6206. https://doi.org/10.1002/2014GL060881
Grieger, J., Leckebusch, G. C., Donat, M. G., Schuster, M., & Ulbrich, U. (2014). Southern Hemisphere winter cyclone activity under recent
and future climate conditions in multi-model AOGCM simulations. International Journal of Climatology, 34(12), 3400–3416. https://doi.
org/10.1002/joc.3917
Grieger, J., Leckebusch, G. C., Raible, C. C., Rudeva, I., & Simmonds, I. (2018). Subantarctic cyclones identified by 14 tracking methods, and
their role for moisture transports into the continent. Tellus, Series A: Dynamic Meteorology and Oceanography, 70(1), 1–18. https://doi.org/1
0.1080/16000870.2018.1454808
Hobbs, W. R., Massom, R., Stammerjohn, S., Reid, P., Williams, G., & Meier, W. (2016). A review of recent changes in Southern Ocean sea ice,
their drivers and forcings. Global and Planetary Change, 143, 228–250. https://doi.org/10.1016/j.gloplacha.2016.06.008
Hoskins, B. J., & Hodges, K. I. (2005). A new perspective on Southern Hemisphere storm tracks. Journal of Climate, 18(20), 4108–4129. https://
doi.org/10.1175/JCLI3570.1
Keable, M., Simmonds, I., & Keay, K. (2002). Distribution and temporal variability of 500 hPa cyclone characteristics in the Southern Hemi-
sphere. International Journal of Climatology, 22(2), 131–150. https://doi.org/10.1002/joc.728
Krinner, G., Magand, O., Simmonds, I., Genthon, C., & Dufresne, J. L. (2007). Simulated Antarctic precipitation and surface mass balance at the
end of the twentieth and twenty-first centuries. Climate Dynamics, 28(2–3), 215–230. https://doi.org/10.1007/s00382-006-0177-x
Kwok, R., Pang, S. S., & Kacimi, S. (2017). Sea ice drift in the Southern Ocean: Regional patterns, variability, and trends. Elementa, 5. https://
doi.org/10.1525/elementa.226
Lecomte, O., Goosse, H., Fichefet, T., Holland, P. R., Uotila, P., Zunz, V., & Kimura, N. (2016). Impact of surface wind biases on the Antarctic
sea ice concentration budget in climate models. Ocean Modelling, 105, 60–70. https://doi.org/10.1016/j.ocemod.2016.08.001
Lim, E. P., & Simmonds, I. (2002). Explosive cyclone development in the Southern Hemisphere and a comparison with Northern Hemisphere
events. American Meteorological Society, 130, 2188–2209. https://doi.org/10.1175/1520-0493(2002)130<2188:ecdits>2.0.co;2
Lim, E. P., & Simmonds, I. (2007). Southern hemisphere winter extratropical cyclone characteristics and vertical organization observed with the
ERA-40 data in 1979–2001. Journal of Climate, 20(11), 2675–2690. https://doi.org/10.1175/JCLI4135.1
Lynch, A., Uotila, P., & Cassano, J. J. (2006). Changes in synoptic weather patterns in the polar regions in the twentieth and twenty-first centuries
part 2: Antarctic. International Journal of Climatology, 26(9), 1181–1199. https://doi.org/10.1002/joc.1305
Massonnet, F., Fichefet, T., Goosse, H., Vancoppenolle, M., Mathiot, P., & König Beatty, C. (2011). On the influence of model physics on simu-
lations of Arctic and Antarctic sea ice. The Cryosphere, 5(3), 687–699. https://doi.org/10.5194/tc-5-687-2011
Matear, R. J., O'Kane, T. J., Risbey, J. S., & Chamberlain, M. (2015). Sources of heterogeneous variability and trends in Antarctic sea-ice. Nature
Communications, 6, 8656. https://doi.org/10.1038/ncomms9656
McDonald, A. J., & Cairns, L. H. (2020). A new method to evaluate reanalyses using synoptic patterns: An example application in the Ross Sea/
Ross Ice Shelf region. Earth and Space Science, 7(1). https://doi.org/10.1029/2019EA000794
Meier, W. N., & Stroeve, J. (2008). Comparison of sea-ice extent and ice-edge location estimates from passive microwave and enhanced-resolu-
tion scatterometer data. Annals of Glaciology, 43, 65–70. https://doi.org/10.3189/172756408784700743
Messmer, M., & Simmonds, I. (2021). Global analysis of cyclone-induced compound precipitation and wind extreme events. Weather and
Climate Extremes, 32, 100324. https://doi.org/10.1016/j.wace.2021.100324
Nash, D., Waliser, D., Guan, B., Ye, H., & Ralph, M. F. (2018). The role of atmospheric rivers in extratropical and polar hydroclimate. Journal
of Geophysical Research: Atmospheres, 123(13), 6804–6821. https://doi.org/10.1029/2017JD028130
Neu, U., Akperov, M. G., Bellenbaum, N., Benestad, R., Blender, R., Caballero, R., et al. (2013). IMILAST: A community effort to inter-
compare extratropical cyclone detection and tracking algorithms. Bulletin of the American Meteorological Society, 94, 529–547. https://doi.
org/10.1175/BAMS-D-11-00154.1
Nieman, P. J., Ralph, F. M., Wick, G. A., Lundquist, J. D., & Dettinger, M. D. (2008). Meteorological characteristics and overall precipitation
impacts of atmospheric rivers affecting the West Coast of North America based on eight years of SSM/I Satellite observations. Journal of
Hydrometeorology, 9, 22–47. https://doi.org/10.1175/2007JHM855.1
Papritz, L., Pfahl, S., Rudeva, I., Simmonds, I., Sodemann, H., & Wernli, H. (2014). The role of extratropical cyclones and fronts for Southern
Ocean freshwater fluxes. Journal of Climate, 27(16), 6205–6224. https://doi.org/10.1175/JCLI-D-13-00409.1
Patoux, J., Yuan, X., & Li, C. (2009). Satellite-based midlatitude cyclone statistics over the Southern Ocean: 1. Scatterometer-derived pressure
fields and storm tracking. Journal of Geophysical Research, 114, D04105. https://doi.org/10.1029/2008JD010873
Pinto, J. G., Spangehl, T., Ulbrich, U., & Speth, P. (2005). Sensitivities of a cyclone detection and tracking algorithm: Individual tracks and
climatology. Meteorologische Zeitschrift, 14(6), 823–838. https://doi.org/10.1127/0941-2948/2005/0068
Raphael, M. N. (2007). The influence of atmospheric zonal wave three on Antarctic sea ice variability. Journal of Geophysical Research, 112,
D12112. https://doi.org/10.1029/2006JD007852
Simmonds, I. (2000). Size changes over the life of sea level cyclones in the NCEP reanalysis. Monthly Weather Review, 128(12), 4118–4125.
https://doi.org/10.1175/1520-0493(2000)129<4118:SCOTLO>2.0.CO;2
Simmonds, I. (2003). Modes of atmospheric variability over the Southern Ocean. Journal of Geophysical Research, 108(4). https://doi.
org/10.1029/2000jc000542
Simmonds, I. (2015). Comparing and contrasting the behaviour of Arctic and Antarctic sea ice over the 35 year period 1979-2013. Annals of
Glaciology, 56(69), 18–28. https://doi.org/10.3189/2015AoG69A909
Simmonds, I., & Keay, K. (2000a). Mean southern hemisphere extratropical cyclone behavior in the 40-year NCEP-NCAR reanalysis. Journal of
Climate, 13(5), 873–885. https://doi.org/10.1175/1520-0442(2000)013<0873:MSHECB>2.0.CO;2
Simmonds, I., & Keay, K. (2000b). Variability of Southern Hemisphere extratropical cyclone behavior, 1958–97. Journal of Climate, 13(3),
550–561. https://doi.org/10.1175/1520-0442(2000)013<0550:VOSHEC>2.0.CO;2
Simmonds, I., Keay, K., & Bye, J. A. T. (2012). Identification and climatology of Southern Hemisphere mobile fronts in a modern reanalysis.
Journal of Climate, 25(6), 1945–1962. https://doi.org/10.1175/JCLI-D-11-00100.1
Simmonds, I., Keay, K., & Lim, E. P. (2003). Synoptic activity in the seas around Antarctica. Monthly Weather Review, 131(2), 272–288. https://
doi.org/10.1175/1520-0493(2003)131<0272:SAITSA>2.0.CO;2
Simmonds, I., & Li, M. (2021). Trends and variability in polar sea ice, global atmospheric circulations, and baroclinicity. Annals of the New York
Academy of Sciences, 1504, 167–186. https://doi.org/10.1111/nyas.14673
Simmonds, I., & Wu, X. (1993). Cyclone behaviour response to changes in winter Southern Hemisphere sea-ice concentration. Quarterly Journal
of the Royal Meteorological Society, 119, 1121–1148. https://doi.org/10.1002/qj.49711951313
Sinclair, V. A., & Dacre, H. F. (2019). Which extratropical cyclones contribute most to the transport of moisture in the Southern Hemisphere?
Journal of Geophysical Research: Atmospheres, 124(5), 2525–2545. https://doi.org/10.1029/2018JD028766
Smith, G. C., Allard, R., Babin, M., Bertino, L., Chevallier, M., Corlett, G., et al. (2019). Polar ocean observations: A critical gap in the observ-
ing system and its effect on environmental predictions from hours to a season. Frontiers in Marine Science, 6, 429. https://doi.org/10.3389/
fmars.2019.00429
Stuecker, M. F., Bltz, C. M., & Armour, K. C. (2017). Conditions leading to the unprecedented low Antarctic sea ice extent during the 2016 austral
spring season. Geophysical Research Letters, 44, 9008–9019. https://doi.org/10.1002/2017GL074691
Turner, J. (2004). The El Nino-Southern Oscillation and Antarctica. International Journal of Climatology, 24, 1–31. https://doi.org/10.1002/
joc.965
Uotila, P., Holland, P. R., Vihma, T., Marsland, S. J., & Kimura, N. (2014). Is realistic Antarctic sea-ice extent in climate models the result of
excessive ice drift? Ocean Modelling, 79, 33–42. https://doi.org/10.1016/j.ocemod.2014.04.004
Uotila, P., Vihma, T., Pezza, A. B., Simmonds, I., Keay, K., & Lynch, A. H. (2011). Relationships between Antarctic cyclones and surface
conditions as derived from high-resolution numerical weather prediction data. Journal of Geophysical Research, 116(7), 1–14. https://doi.
org/10.1029/2010JD015358
Vichi, M., Eayrs, C., Alberello, A., Bekker, A., Bennetts, L., Holland, D., et al. (2019). Effects of an explosive polar cyclone crossing the Antarctic
marginal ice zone. Geophysical Research Letters, 46(11), 5948–5958. https://doi.org/10.1029/2019GL082457
Wille, J. D., Favier, V., Dufour, A., Gorodetskaya, I. V., Turner, J., Agosta, C., & Codron, F. (2019). West Antarctic surface melt triggered by
atmospheric rivers. Nature Geoscience, 12(11), 911–916. https://doi.org/10.1038/s41561-019-0460-1
Wille, J. D., Favier, V., Gorodetskaya, I. V., Agosta, C., Kittel, C., Chowdhry Beeman, J., et al. (2021). Antarctic atmospheric river climatology
and precipitation impacts. Journal of Geophysical Research: Atmospheres, 126, ee2020JD033788. https://doi.org/10.1029/2020JD033788
Xu, G., Ma, X., Chang, P., & Wang, L. (2020). Image-processing-based atmospheric river tracking method version 1 (IPART-1). Geoscientific
Model Development, 13, 4639–4662. https://doi.org/10.5194/gmd-13-4639-2020
Yuan, X., Patoux, J., & Li, C. (2009). Satellite-based midlatitude cyclone statistics over the Southern Ocean: 2. Tracks and surface fluxes. Journal
of Geophysical Research, 114, D04106. https://doi.org/10.1029/2008JD010874
Zhang, Z., Ralph, F. M., & Zheng, M. (2019). The relationship between extratropical cyclone Strength and atmospheric river Intensity and posi-
tion. Geophysical Research Letters, 46(3), 1814–1823. https://doi.org/10.1029/2018GL079071
Zhu, Y., & Newell, R. E. (1998). A proposed algorithm for moisture fluxes from atmospheric rivers. Monthly Weather Review, 126(3), 725–735.
https://doi.org/10.1175/1520-0493(1998)126<0725:APAFMF>2.0.CO;2

Paper 8 Imp

Uploaded by

Copyright:

Available Formats

Paper 8 Imp

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Paper 8 Imp

Uploaded by

Copyright:

Available Formats

RESEARCH ARTICLE Association Between Extreme Atmospheric Anomalies Over

Received 4 NOV 2021

2. Data and Methods

2.2. Atmospheric Anomalies

2.3. Detection of Cyclone and Atmospheric River Tracks

2.4. Search-Radius Method and Random Association

3.2. Distribution of Synoptic Features and Extreme Atmospheric Anomalies

4. Discussion and Conclusions

Data Availability Statement

You might also like