- Annick TerpstraAnnick TerpstraUniversity of Bergen
- and Shun-ichi WatanabeShun-ichi WatanabeMeteorological Research Institute, Japan Meterological Agency
Polar lows are intense maritime mesoscale cyclones developing in both hemispheres poleward of the main polar front. These rapidly developing severe storms are accompanied by strong winds, heavy precipitation (hail and snow), and rough sea states. Polar lows can have significant socio-economic impact by disrupting human activities in the maritime polar regions, such as tourism, fisheries, transportation, research activities, and exploration of natural resources. Upon landfall, they quickly decay, but their blustery winds and substantial snowfall affect the local communities in coastal regions, resulting in airport-closure, transportation breakdown and increased avalanche risk.
Polar lows are primarily a winter phenomenon and tend to develop during excursions of polar air masses, originating from ice-covered areas, over the adjacent open ocean. These so-called cold-air outbreaks are driven by the synoptic scale atmospheric configuration, and polar lows usually develop along air-mass boundaries associated with these cold-air outbreaks. Local orographic features and the sea-ice configuration also play prominent roles in pre-conditioning the environment for polar low development. Proposed dynamical pathways for polar low development include moist baroclinic instability, symmetric convective instability, and frontal instability, but verification of these mechanisms is limited due to sparse observations and insufficient resolution of reanalysis data.
Maritime areas with a frequent polar low presence are climatologically important regions for the global ocean circulation, hence local changes in energy exchange between the atmosphere and ocean in these regions potentially impacts the global climate system. Recent research indicates that the enhanced heat and momentum exchange by mesoscale cyclones likely has a pronounced impact on ocean heat transport by triggering deep water formation in the ocean and by modifying horizontal mixing in the atmosphere. Since the beginning of the satellite-era a steady decline of sea-ice cover in the Northern Hemisphere has expanded the ice-free polar regions, and thus the areas for polar low development, yet the number of polar lows is projected to decline under future climate scenarios.
The “Polar Low” Earmark
Vigorous storms are ordinary events over high latitude maritime regions in both hemispheres. These storms usually last for about a day and are estimated to be several hundred kilometers in diameter, though occasionally they can last up to two days and attain sizes up to 1,000 km. The storms are most prevalent during the local winter season and usually develop concurrent with excursions of cold polar air-masses into lower latitudes. Strong wind speeds associated with these storms, sometimes reaching the lower limit of the hurricane wind-scale (Saffire-Simpson), result in local enhancement of heat, moisture, and momentum exchange between the ocean and atmosphere. These wind-driven interactions can result in rough sea states and, combined with cold temperatures, provide challenging conditions for local fisherman and other maritime activities. Some of these predominantly maritime storms progress into coastal regions, where they often produce significant snowfall within a timespan of several hours. The most common name for these intense, short-lived, relatively small winter storms is polar low.
Characterizing what exactly entails a polar low has been a topic of debate in the past, and even today, there is no unambiguous definition of a polar low. Factors contributing to this discussion are the lack of consensus on a conceptual model for polar lows, the lack of clearly defined thresholds for selection criteria, the broad spectrum and transitional nature of dynamical forcing mechanisms, and global differences in their appearance. Suggested definitions of polar lows incorporate various restrictions on size, intensity, duration, cloud-structure, propagation direction, etc., (e.g., Carleton, 1985; Rasmussen & Turner, 2003; Reed & Duncan, 1987; Zahn & von Storch, 2008). However, most studies tend to agree on two distinctive features: polar lows are intense maritime mesoscale cyclones, and they develop on the cold side of the mid-latitude jet-stream. As such they can be considered a subcategory of all polar mesoscale vortices, constituting only the most intense polar mesoscale cyclones.
In the early days of polar low research, infrared satellite images were used to categorize polar lows based on the appearance of their cloud-structures (see Figure 1). The major distinguishing feature used in these studies was between comma-shaped and spiraliform cloud features, where the latter is rather uncommon, as only a few polar lows exhibit hurricane-like, spiraliform cloud structures (Mokhov, Akperov, Lagun, & Lutsenko, 2007). This type of classification and the lack of more conventional observations nurtured the debate, in the 1980s and early 1990s, about the convective versus baroclinic nature of polar lows, including coining the term Arctic hurricane for polar lows (Emanuel & Rotunno, 1989). Categorization of polar lows based on dynamical paradigms is limited, as forcing mechanisms for polar lows cover a broad range of mechanisms, and polar lows frequently transition between dominant mechanisms during the genesis, intensification, and maintenance phases of their development. Another approach to categorize polar lows is by considering the characteristics of their genesis environment, where Businger and Reed (1989) distinguished three basic configurations: (a) short-wave/jet-streak type, characterized by a mobile short-wave trough at upper levels triggering cyclogenesis, (b) arctic front type, associated with pre-existing low-level baroclinic frontal structures, and (c) cold-low type, characterized by strong surface fluxes and deep convection inside the cold air-mass. Although they identified distinctive polar low genesis environments, it remains unclear which of these categories are prevalent and their relevance for polar low genesis. Several other categorization schemes (e.g., Grønås & Kvamstø, 1995), mainly variations of the above, were proposed, but none of them pinpointed the actual frequency of occurrence or their relevance for polar low formation. A more recently developed classification of ambient polar low genesis environments is based on the angle between the mean wind and the thermal wind vector in the lower troposphere, thus distinguishing between two dynamically different genesis environments: (1) forward-shear, comprising tropospheric deep baroclinicity, similar to environments associated with classical midlatitude cyclogenesis, and (2) reverse-shear, associated with an intense baroclinic zone confined to the lower troposphere in an otherwise barotropic troposphere. Examination of the genesis environments of polar low in an existing polar low database (STARS, Sætra et al., 2010) shows that both environments equally constitute polar low genesis (Terpstra, Michel, & Spengler, 2016), though not all polar lows develop during strictly forward/reverse shear conditions. Considering objectively identified polar mesoscale cyclones in reanalysis data indicate that reverse-shear environments are less common among mesoscale cyclones, but are associated with the most intense mesoscale cyclones (Michel, Terpstra, & Spengler, 2018).
As polar lows comprise the most intense polar mesoscale cyclones, common criteria to distinguish polar lows from other cyclonic features often include a minimum threshold for low-level wind speed. Values for this threshold are generally larger than 15 m s−1; however, the exact values depend on the purpose of the study and the origin on the data—for example, observations demand different thresholds than output from climate models. Another frequently used criteria is the requirement of low static stability in the troposphere, usually prescribed by the difference between the sea surface temperature (SST) and the air temperature at 500 hPa. Values for this stability threshold vary regionally due to climate scale differences in polar low regions: for example, Pacific polar lows tend to develop in more stable regions than their Atlantic counterparts. Furthermore, this criterion favors the selection of certain types of polar lows as more convective, and shear-line polar lows tend to develop in less stable conditions (Michel et al., 2018; Sergeev, Renfrew, Spengler, & Dorling, 2017; Smirnova & Golubkin, 2017; Terpstra et al., 2016). Applying the latter could have implications for assessing future changes in polar lows in climate models, as it is unclear if a reduction in one type of polar lows is compensated by another type. Additional restrictions such as propagation direction, cold-air outbreak indices, and spatial limitations are often used in climatologies of polar lows, with some studies examining the effects of various restrictions on the selection of features (Bracegirdle & Gray, 2008; Michel et al., 2018; Stoll, Graversen, Noer, & Hodges, 2018).
The low spatial density of surface-based observations in maritime high-latitude regions combined with the short lifespan and rather small horizontal scales of polar mesoscale cyclones, contribute to polar lows often not being captured by the conventional observing network. To bypass this lack in observations, early investigations of polar lows relied on visible and infrared satellite images to analyze their cloud structure and spatial distribution (e.g., Asai, 1988; Blechschmidt, 2008; Carleton, 1985; Mokhov et al., 2007). In recent years, the growing volume of satellite products has contributed to an increase in observational data for exploring properties of polar lows. For example, NASA CloudSat satellite, with a 94GHz cloud radar, provides a nadir view with 240m horizontal resolution with coverage in polar region. A variety of derived science products, such as cloud top height and liquid and ice water content, have been used to validate numerical modeling studied (Sergeev et al., 2017) and to study the properties of clouds associated with polar lows (Forsythe & Haynes, 2015). As most polar lows exhibit a warm core relative to their ambient environmental, they are also clearly detectable using brightness temperature differences derived from space-borne passive microwave sounders (Moore & Vonder Haar, 2003), or hydrometeors in clouds (Claud, Funatsu, Noer, & Chaboureau, 2009; Melsheimer, Frost, & Heygster, 2016). The structure and intensity of the wind field imprinted on the ocean by polar lows can be estimated with scattometer and Synthetic Aperture Radar (Moore & Vachon, 2002; Song, Subrahmanyam, Guo, & Panda, 2015). Polar lows are associated with enhanced atmospheric moisture content with respect to their environment, which enables clear detection of their structures by Special Sensor Microwave Imager (SSM/I) and Advanced Microwave Scanning Radiometer-Earth Observing System (AMSR-E) (Bobylev, Zablotski, Mitnik, & Mitnik, 2011; Gurvich, 2013)
In-situ observations of polar lows are rare, partly due to expensive and complicated logistical issues in the harsh maritime polar regions in winter and partly due to difficulties in accurate prediction of polar lows as a result of their short lifespan and small scales. Hence, the handful of targeted airborne observations of polar lows provide valuable insight in the structures associated with polar lows (see Figure 2 for an example), and include airborne observations, dropsondes, and more recently, airborne lidar measurements. Most of the targeted polar lows (Bond & Shapiro, 1991; Brümmer, Müller, & Noer, 2009; Douglas, Shapiro, Fedor, & Saukkonen, 1995; Føre et al., 2011; Murakami, 2019; Sergeev et al., 2017; Shapiro, Fedor, & Hampel, 1987; Wagner, Gohm, Dörnbrack, & Schäfler, 2011) were in their mature stage. They exhibited horizontal scales between 300 and 500 km with relatively shallow cloud-top heights (around 3 km), although occasionally cloud tops reached 7 km. Maximum values of specific humidity in the observed polar lows ranged between 2–4 g kg−1, and estimated precipitation, available for only a few cases, was around 0.3–1.5 mm h−1. Observed low-level maximum wind speeds ranged between 20–35 m s−1, and estimated surface sensible and latent heat fluxes varied between 200–500 W m−2 each. Structurally, these polar lows featured a core that was 3–5 K warmer than their environment, and often these polar lows were accompanied by frontal structures similar to mid-latitude extratropical cyclones.
Climatologies of Polar Lows
Frequent polar low genesis takes place in both the Arctic and Antarctic ice-free regions, but so far the bulk of polar low research has focused on the North Atlantic region. Prominent regions for polar low development in the Northern Hemisphere include the Nordic Seas (e.g., Blechschmidt, 2008; Ese, Kanestrom, & Pedersen, 1988; Michel et al., 2018; Noer, Saetre, Lien, & Gusdal, 2011), Japan Sea (e.g., Watanabe, Niino, & Yanase, 2016; Yanase, Niino, & Watanabe, 2016), Labrador Sea (e.g., Mallet, Claud, Cassou, Noer, & Kodera, 2013; Pagowski & Moore, 2001) and the Gulf of Alaska (e.g., Bond & Shapiro, 1991; Businger, 1987). In the Southern Hemisphere, polar lows are frequently found around the Amundsen and Bellingshausen Sea and off the coast of Wilkes Land, Australia, (e.g., Carleton & Song, 1997; Irving, Simmonds, & Keay, 2011; Pezza et al., 2015). Other common polar low regions in the Southern Hemisphere include the Ross Sea and Weddell Sea (e.g., Bromwich, 1991; Carrasco & Bromwich, 1993; Heinemann, 1990; Verezemskaya, Tilinaina, Gulev, Renfrew, & Lazzara, 2017).
Climatologies of polar lows are based on satellite images (e.g., Noer et al., 2011; Wilhelmsen, 1985), reanalysis data (e.g. Bracegirdle & Gray, 2008; Condron, Bigg, & Renfrew, 2006; Michel et al., 2018; Stoll et al., 2018; Zappa & Shaffrey, 2014) and dynamical downscaling (e.g. Chen & von Storch, 2013; Zahn & von Storch, 2008). Depending on the available data, automated tracking or manual detection is used to identify polar lows. Given their small horizontal scales polar lows are often poorly represented in reanalysis datasets due to rather coarse grid spacing, as smaller and weaker phenomena remain unresolved (Condron et al., 2006; Laffineur, Claud, Chaboureau, & Noer, 2014; Smirnova & Golubkin, 2017; Zappa & Shaffrey, 2014). To circumvent this issue, several climatologies have resorted to potentially favorable conditions for polar low development such as cold-air-outbreaks, reduced static stability, and reverse shear conditions (e.g., Claud, Duchiron, & Terray, 2007; Kolstad, 2006, 2011; Landgren et al., 2019), as opposed to identifying actual polar lows to compile their climatologies. Observations in high-latitude maritime regions are still sparse compared to lower latitudes, further limiting verification of these polar low climatologies. Providing the above mentioned complications for compiling polar low climatologies and the range of choices of selection criteria, climatologies are inconclusive about the number of annual events, with estimates ranging between 15 and 150 polar low cases per year for individual ocean basins. However, most climatologies seem to roughly agree on the spatial distribution of areas of frequent polar low formation (see Figure 3).
Favorable Genesis Conditions
Synoptic-Scale Configuration: Cold Air Outbreak
Intensification of incipient polar cyclonic disturbances depends on their location relative to the synoptic-scale flow (Forbes & Lottes, 1985), with polar low genesis typically taking place in the rear of an upper-level trough (Blechschmidt, 2008; Businger, 1985; Mallet et al., 2013). This configuration causes large scale equatorward geostrophic flow, often accompanied by cyclonic vorticity advection, resulting in advection of cold air from ice-covered regions over the adjacent open ocean. The accompanying cold temperature anomaly at upper levels reduces the static stability over the polar low genesis region. The build-up period prior to this favorable synoptic scale setup is approximately 4 days, in which the associated height and temperature anomalies intensify over the genesis area, with a maximum just before the genesis period (Blechschmidt, 2008; Businger, 1985; Mallet et al., 2013; Yanase, Niino, & Watanabe, 2016). The flow associated with the synoptic scale low to the east of the genesis location encourages advection of cold air masses originating from ice-covered areas into the ice-free regions. These so-called cold-air outbreaks usually last for several days, although occasionally persistent meandering of the jet-stream (blocking) induces longer lasting cold air outbreak events. Due to the large air-sea temperature contrast, cold air outbreaks are accompanied by relative strong surface fluxes, and account for a significant portion of the total atmosphere-ocean heat exchange at high latitudes (Papritz & Spengler, 2017). At the periphery of these cold-air outbreaks, cold dry continental and warm moist maritime air masses meet, making these fringes favorable regions for mesoscale cyclone development. Polar low development is typically associated with these cold-air outbreaks, and it is not uncommon that several mesoscale cyclones develop during a single cold-air outbreak event.
Pre-Conditioning By Local Features
As the synoptic scale configuration sets the scene for polar low development by inducing a cold air outbreak and reducing the static stability over the region, local effects associated with these excursions of cold air masses also contribute to the potential for development of polar lows. First, the combination of extremely cold air over a relative warm ocean introduces strong surface fluxes, which assist in decreasing the static stability, deepening and hydrating the boundary layer (e.g. Hartmann, Kottmeier, & Raasch, 1997). Hence, as the CAO progresses, the cold air mass develops an unstable stratification below, which favors the potential for cyclone development. Second, there are several sources of local low-level vorticity during cold air outbreak events that can support the initiation of mesoscale cyclogenesis. One source of vorticity is the ice-edge, characterized by strong gradients in moisture, temperature, and roughness. Differential heating from the surface along the ice-edge, or even along strong sea-surface temperature gradients, can introduce low-level baroclinic zones that subsequently are advected away from their source. Occasionally, polar lows develop along such ice-edge induced low-level baroclinic zones (e.g., Douglas et al., 1995; Heinemann, 1996; Shapiro & Fedor, 1989). In addition, polar low genesis regions often include pre-existing fronts or cloud bands associated with the synoptic scale configuration. Initiation of polar lows along such antecedent frontal structures are commonly referred to as reverse shear polar low development (for an example, see Bond & Shapiro, 1991; Figure 4). Regions of frequent polar low development are often surrounded by significant orographic features, such as Greenland and Svalbard in the Nordic Seas area, Korea in the Japan Sea region and the Antarctic continent in the Southern Hemisphere, and/or an ice-edge or a coastline with characteristic shapes. Interaction of the flow with local orography or sharp bends in the ice-edge can induce low-level vorticity sources including shear-lines and convergence zones, which subsequently initiate polar low development (e.g. Martin & Moore, 2006; Mingalev, Orlov, & Mingalev, 2014; Moore & Vachon, 2002; Sergeev, Renfrew, & Spengler, 2018; and see Figure 5 for an example). A well-known persistent wintertime convergence zone along which polar lows develop is the Japan Sea Polar air-mass Convergence Zone (JPCZ; Asai, 1988), which emerges during cold-air outbreaks due to the land-sea thermal contrast and blocking of the flow by the mountains North of Korea (Nagata, 1991; Watanabe & Niino, 2014). In the Antarctic region, katabatic winds off the continent also provide a source of localized convergence zones, which potentially provide suitable conditions for mesoscale cyclogenesis (Bromwich, 1991; Klein & Heinemann, 2002).
The dynamical pathways underlying polar low formation are still under development. The continuous controversy with regard to their dominant forcing mechanisms contributed to the notion of a “spectrum of polar lows,” where the paradigms of “baroclinic instability” and “symmetric convection” represent the two ends of this spectrum; yet other mechanisms, such as frontal instability, also remain plausible candidates. Furthermore it is not uncommon for polar lows to transition between different forcing mechanisms during their life-cycle, rendering it difficult to associate a single dynamical mechanism to their development. Due to the lack of observations to validate models, there is still a discrepancy between proposed theories and actual polar low development.
Some of the early investigations of polar lows suggested baroclinic instability, a theoretical concept developed to describe the wave-like patterns of mid-latitude cyclones and anticyclones (Charney, 1947; Eady, 1949), as the dominant forcing mechanism. During baroclinic instability, the energy required for perturbation growth is supplied by an uneven distribution of potential energy in the troposphere, where the simultaneous ascent of warm air and descent of cold air results in conversion of this reservoir of potential energy into kinetic energy. Based on simple dry linear models for evaluating the applicability of this concept to polar lows, it turned out that dry baroclinic instability theory was not very successful in predicting polar low development (Duncan, 1977; Mansfield, 1974; Reed & Duncan, 1987). The paradigm was unable to explain the short-length scales and rapid evolution of polar lows, and most studies attribute this mismatch to the lack of moisture in their analysis. The inclusion of moisture slightly reduces the static stability of the atmosphere, however the impact of moisture on baroclinic instability is most prominent during condensation or evaporation. The latent heating associated with phase transitions (e.g., cloud/precipitation formation) introduce local heating or cooling thereby redistributing the available potential energy, typically resulting in enhanced growth rates and reduced wavelengths compared to their dry counterpart. The importance of latent heating for polar low development is underscored in numerous numerical case studies where turning off the latent heating resulted in reduced amplitude for the polar lows (e.g., Bresch et al., 1997; Claud, Heinemann, Raustein, & McMurdie; 2004; McInnes Kristiansen, Kristjansson, & Schyberg, 2011; Nordeng & Rasmussen, 1992).
Latent heating due to condensation is primarily achieved by lifting of air, where the distribution of the latent heating depends on the lifting mechanism: slantwise ascent related to baroclinic processes tends to spread the heating horizontally, whereas deep convection tends to increase the depth of the latent heating. Nuances in the distribution of latent heating determine the effect on cyclone intensification. Focusing on polar lows, Craig and Cho (1988) explored effects of the depth and intensity of latent heating using an idealized numerical framework. They identified a threshold value for latent heating to become effective in modifying baroclinic polar low development, the value of this threshold depends on the vertical distribution of the latent heating. Furthermore, low values of latent heating resulted in destabilization, which enhanced the growth rates, whereas high values of latent heating resulted in latent heating dominating the conversion of available potential energy into kinetic energy. The effectiveness of latent heating on perturbation growth depends on the vertical gradients of latent heating, as opposed to absolute values. Hence, low absolute values of latent heating at high latitudes due to limited moisture availability are potentially compensated by the reduced vertical extent of cyclones at high latitudes (Terpstra, Spengler, & Moore, 2015).
Recognition of the prominent role of latent heating in polar lows prompted the suggestion of several variations of moist baroclinic instability theory as potential dynamical pathways for polar low formation. Montgomery and Farrell (1992) developed a conceptual model for polar low formation in which the modification of the initially dry baroclinic evolution by latent heat release is sufficient to generate polar lows, and Bracegirdle and Gray (2008) estimated that about 30% of polar lows develop similar to type-C cyclogenesis, a concept introduced by Deveson, Browning, and Hewson (2002) for moist baroclinic mid-latitude cyclogenesis in which the intensification stage is dominated by latent heating. Another potential moist baroclinic pathway for polar low formation entails the so-called Diabatic Rossby Waves, a concept put forward by Snyder and Lindzen (1991) and Parker and Thorpe (1995), in which the latent heating linked solely to a surface-based cyclonic circulation acts as the dominant forcing for cyclone intensification. This mechanism is fundamentally different from classical baroclinic instability, and idealized simulations of polar low development underpin the potential of this pathway for polar low development (Terpstra et al., 2016).
Another way to conceptualize baroclinic instability is via interactions between confined areas of anomalous potential vorticity (PV), in which a positive feedback between an upper level (tropopause) and low-level PV anomaly result in intensification of the cyclone (Hoskins, McIntyre, & Robertson, 1985). In an adiabatic, frictionless flow, PV fields can be directly associated with the temperature and velocity field, facilitating assessment of the relative contributions of the individual PV anomalies to the overall cyclone intensification. Note that this requires assuming balanced conditions, which are not necessarily valid for mesoscale features, such as polar lows. Applying this principle to polar lows demonstrated the dominant role of a tropopause based PV anomaly for the intensification (Bracegirdle & Gray, 2009), or the during the entire life-cycle of several polar low cases (Nordeng & Røsting 2011; Wu, Martin, & Petty, 2011). Polar low development is frequently preceded by upper level PV anomalies—short-waves propagating at the height of the tropopause, approaching the genesis region (e.g., Claud et al., 2004; Grønås , Foss, & Lystad, 1987; Mailhot, Hanley, Bilodeau, & Hertzman, 1996; Nordeng & Røsting, 2011; Pagowski & Moore, 2001; Rasmussen, 1985; Shimada, Wada, Yamazaki, & Kitabatake, 2014).
Symmetric Convective Instability
Sporadically polar lows exhibit hurricane-like cloud structures in satellite images—spiraliform cloud-bands with a distinct cloud-free eye. This prompted a hypothesized pathway for polar low formation based on symmetric convective theories derived from conceptual models for tropical cyclone formation. One such paradigm is Convective Instability of the Second Kind (CISK), which is based on low-level moisture convergence as a source for latent heat release, where a positive feedback between the latent heating and low-level convergence acts to sustain and intensify the cyclone. The effectiveness of CISK depends on the available amount of convective potential energy (CAPE). Several studies suggested that CISK can be active during polar low formation (e.g. Økland, 1986; Rasmussen, 1979) thereby implying a sufficient reservoir of CAPE. However, recent analysis of dropsondes observation in the vicinity of polar lows indicated that CAPE is virtually absent during polar low formation (Linders & Saetra, 2010). Thus it remains an open question if CISK contributes to the formation of polar lows. Emanuel and Rotunno (1989) and Gray and Craig (1998) used an axisymmetric model to show that at least some polar lows can be driven by Wind Induced Surface Heat Exchange (WISHE). This mechanism underscores the direct relation between surface fluxes and low-level wind speeds, where a positive feedback is established via low-level moisture convergence due to surface fluxes initiating latent heat release, which in turn intensifies the low-level wind speed and thus maintaining and enhancing the surface fluxes. However, they emphasis that an initial sufficiently strong cyclone is a prerequisite for the mechanisms to sustain this positive feedback. Another complicating factor for the applicability of symmetric convective paradigms to polar low development is that these theories assume axisymmetric circulations, whereas the majority of observed polar lows do not exhibit this symmetric property (Blechschmidt, 2008), and if they do they are already in their mature phase.
The Baroclinic-Convective Spectrum
Fantini and Buzzi (1993) conducted one of the first studies exploring the possibility of polar low development in an initial baroclinic atmosphere where surface fluxes drive subsequent transition into more convection dominated cyclone growth and maintenance. In their 2D model, the energy supplied by surface fluxes is initially stored in the planetary boundary layer during the baroclinic phase of the cyclone growth, until the initiation of deep convection transforms the cyclone growth in a more hurricane-like structure. More recently, Yanase and Niino (2007) used idealized 3D numerical simulations to explore the baroclinic-convective parameter space and the role of surface fluxes on polar low development. Depending on the initial baroclinicity and moisture content, the structural evolution of these polar lows resembled hurricane-like, surface flux driven development in the absence of initial baroclinicity, where increasing the baroclinicity resulted in more classical mid-latitude cyclogenesis dominated by moist baroclinicity (see Figure 6). Note, however, that the majority of polar low case studies, if not all, exhibit some form of baroclinicity in their genesis environments. Hence baroclinic instability seems a likely candidate for dynamical forcing during the initial stages of polar low formation, with surface fluxes and latent heating modifying the cyclogenesis process. Convective processes, either as the dominant mechanism or by modifying the baroclinic process likely play a significant role during polar low intensification, with occasionally polar lows transitioning during their life-cycle into more hurricane-like structures.
Atmospheric boundaries at low-levels such as fronts or convergence zones, either embedded in a cold-air-outbreak or at the outer fringes of the cold air masses, are frequently associated with polar low genesis. For example, in the Nordic Seas region a significant fraction of polar lows develop along low-level frontal zones (Michel et al., 2018; Terpstra et al., 2016;). In the Japan Sea region, mesoscale cyclogenesis is often related to the JPCZ (Nagata, 1993; Watanabe & Ninno, 2014; Watanabe, Niino, & Yanase, 2016, 2018), and there are numerous case studies of polar lows that develop in the vicinity of low-level frontal zones (Bond & Shapiro, 1991; Claud et al., 2004; Føre et al., 2011; Grønås, Foss, & Lystad, 1987, Grønås & Kvamstø, 1995; Mailhot et al., 1996; Nordeng & Røsting, 2011; Rasmussen, 1985; Reed & Duncan, 1987; Sergeev et al., 2017).
These frontal structures are usually accompanied by a strong baroclinicity and cyclonic horizontal shear, which resembles the environments associated with frontal instability growth (Joly & Thorpe, 1989; Schar & Davis, 1990), and mesoscale cyclone formation could be triggered by this shear-driven mechanism. However, the relevance of these frontal structures for polar low initiation remains an open question and is case-dependent. In some cases, barotropic energy conversion, in which perturbation growth occurs via conversion of the basic state kinetic energy into eddy kinetic energy, is dominant (e.g., Nagata, 1993), while moist baroclinic processes are dominant in other cases (e.g., Bond & Shapiro, 1991). Recent studies using a convection permitting model show that small vortices that were generated via barotropic instability on a convergence zone were intensified by convection and eventually merged into a polar low (Sergeev et al., 2017; Watanabe & Ninno, 2014).
Another pathway for polar low genesis associated with low-level frontal instability is interaction with an upper-level PV anomaly. A number of case studies demonstrate that polar lows develop when an upper-level PV anomaly approaches a low-level front (Claud et al., 2004; Grønås et al., 1995; Mailhot et al., 1996; Nordeng & Røsting 2011; Pagowski & Moore, 2001; Rasmussen, 1985; Shimada et al., 2014). The circulation associated with this upper-level PV anomaly deforms the low-level front, resulting in cyclogenesis, which resembles the frontal cyclogenesis discussed in Thorncroft and Hoskins (1990). In addition, an migrating upper-level PV anomaly decreases the stratification and enhances convection associated with the front, resulting in more favorable conditions for cyclogenesis.
Another contribution to polar low development includes the role of radiative processes. By using an axisymmetric model, Craig (1995) showed that differences in radiative cooling between the ambient polar low environment and the cooling rates at the relative warm polar low core can enhance growth rates and increase the maximum intensity. However, axisymmetric conditions during polar low development are rare, and generally associated with the mature phase of polar low development, hence radiative processes potentially only act to maintain and enhance the longevity of polar lows. Note that polar lows are primarily a high-latitude winter-time phenomena, they occur during conditions of very low incoming solar radiation with virtually no diurnal cycle.
Spatial Differences in Dynamical Mechanisms
The climatological conditions in the different ocean basins at high latitudes contribute to the diversity in dominant forcing mechanisms in different regions. With respect to convective contributions during polar low development, the relative warm sea surface temperatures of the Nordic Seas contribute to stronger destabilization of the lower troposphere during cold-air outbreaks, and thus ambient environments which are more conducive to convective polar low development compared to the Pacific or Antarctic region in which the ocean currents are more zonal. Pacific polar lows are found much further south than their North Atlantic counterparts (Mullen, 1979). In addition to the influence of the ocean temperature, flow distortion by orographic features in the region can dictate favorable regions for polar low development. Examples of such regions include the JPCZ in the Japan Sea, which acts as the primary polar low breeding ground in this region, and the high density of polar low development downstream of Svalbard in the Norwegian Sea (e.g., Michel et al., 2018). Polar lows in the Southern Hemisphere exhibit a less pronounced annual cycle than their Northern Hemispheric counterparts. These spatial differences render it difficult to compile global climatologies, or to pinpoint the dominant forcing mechanisms for polar low development without taking into account regional climatological differences.
Climate Scale Interactions
Relation to Climate Scale Variability
As polar low genesis tends to occur during preferred synoptic-scale states, their potential for development is likely related to spatial and temporal synoptic-scale variability. Several studies have evaluated the relationship between climate scale atmospheric modes and polar low development in the North Atlantic region. Synoptic activity is reduced over the Norwegian Seas during the negative phase of the North Atlantic Oscillation (NAO) and positive phase of Scandinavian blocking, therefore Claud et al. (2007) concluded that these conditions are less favorable for polar low formation. Mallet et al. (2013) indicated that the majority of polar lows develop during the Atlantic Ridging regime and confirmed that polar low formation occurs more often during NAO+ conditions than during NAO-. However, Michel et al. (2018) considered the 500 hPa height anomalies during mesoscale cyclones over the Nordic Seas, and their dipoles neither resemble NAO nor Scandinavian Blocking, but rather a combination of both patterns. The lack of representing actual weather patterns hints that classical weather regimes might be rather limited as indicators for mesoscale cyclone formation, including polar lows. Furthermore, the different regions (Labrador Sea, Norwegian Sea, Barents Sea) are affected by different flow regimes during a single weather regime, for example Claud et al. (2007) found favorable changes in conditions over the Norwegian Sea during NAO+, but no relevant changes over the adjacent Barents and Kara Sea during NAO+. Polar low genesis over the North Pacific is linked to large-scale pressure patterns, as more polar lows develop where cold air outbreaks are intensified (Chen & von Storch, 2013). Polar low genesis in the central and western North Pacific region is correlated with the Pacific/North American (PNA) pattern and polar low development over the Okhotsk Sea is correlated with Pacific Decadal Oscillation (PDO). In the Japan Sea, there is a correlation between the number of polar mesocyclone occurrence and Arctic Oscillation (AO) or PDO, with enhanced mesoscale cyclone formation during AO- over the entire Japan Sea and during PDO+ over the northeastern Japan Sea.
Polar lows over the Nordic Seas exhibit a strong inter-annual variability, though there are no long-term trends in polar low frequency over the last few decades (Michel et al., 2018; Stoll et al., 2018; Zahn & von Storch, 2008). The same is true for the North Pacific including the Japan Sea and Okhotsk Sea (Chen & von Storch, 2013). With respect to future climate scenarios, Zahn and von Storch (2010) identified a decrease in North Atlantic polar lows over the next century based on data obtained from dynamically downscaling climate model output. These results are in agreement with polar lows identified in a wide range of climate models (CMIP5), where most of the models (>60%) indicate an overall decrease in polar lows for the RCP8.5 (most severe) climate scenario (Romero & Emanuel, 2017). Note that both studies, by imposing their selection criteria for polar lows, emphasize the convective nature of polar lows, and thereby likely restrict their analysis to a particular type of polar lows. The amplitude of the response to climate change at high-latitudes is more pronounced in the atmosphere than in the ocean—the atmosphere is warming more rapid than the ocean. Hence, destabilization of the lower atmosphere, which depends on the air-sea temperature difference, is likely to decrease in future scenarios, providing a plausible explanation for a reduction in polar lows dominated by convection. However, a significant number of polar lows in current climate conditions develop in more stratified conditions ( Smirnova & Golubkin, 2017; Terpstra et al., 2016). Thus, it remains unclear if the reduction in convective polar lows in future scenarios is compensated by an increase in more baroclinic polar lows.
Interplay Between Polar Lows and the Global Climate System
Both high-latitude areas, and thus polar low regions, have experienced rapid environmental changes during the last decades, including atmospheric warming and enhanced melting of the ice-caps, which could be relevant to polar low development. Polar lows are one of the components in those environments and several studies suggest links between local mesoscale atmospheric features, such as polar lows, and the larger climate system.
The high-latitude North Atlantic region is an area crucial for deep water formation and experiences frequent cold air outbreaks and polar low development. Due to their small scales and short life-times polar mesoscale cyclones are often lacking or too weak in climate models (Condron et al., 2006). By forcing an ocean-only model with amplified wind and thermal forcing representative of mesoscale cyclones, Condron and Renfrew (2013) showed that the induced momentum and heat fluxes, and thus mesoscale cyclones, can have a substantial impact on the oceanic heat transport, thereby influencing the Atlantic meridional overturning circulation. Papritz and Pfahl (2016) considered a case study of polar lows in the Southern Hemisphere and concluded that polar lows play an role in the erosion of cold air masses via warming of the cold-air due to latent heat release associated with the polar low. This study indicates that the under-representation of mesoscale cyclones in climate models could potentially bias the life-time of cold air outbreaks, thereby altering the effects of cold air outbreak induced air-sea exchange. These studies demonstrate that the local effects of polar mesoscale cyclones, either directly via enhanced momentum and heat exchange or indirectly via modification of the cold air masses can have significant impact on the ocean heat transport and thereby impact the global climate system.
Despite the decline in sea-ice extent during the last decades, there are no significant trends in polar low frequency during this period (Michel et al., 2018; Zahn & von Storch, 2008). With regard to future climate simulations, the spatial distribution of polar lows shifts under retreating sea-ice conditions in in these simulations (Romero & Emanuel, 2017; Zahn & von Storch, 2010). The reason for these changes are unclear, but are potentially related to changes in the overall patterns of atmospheric stability, extension of the ice-free regions, or changes in shear-line formation by modifications in the shape and location of the ice-edge.
Overall, research suggests that polar mesoscale cyclones could play a role in modifying the high-latitude environment, and thus indirectly the entire climate system, though how much climate changes affect future polar low development and the role of polar lows in the larger climate system remain open questions.
Numerical Modeling And Forecasting
Due to the sparseness of conventional observations, polar lows are usually first observed in satellite images. However, in polar low regions with significant human activity, such as the Japanese and the Norwegian coastal areas, the local meteorological services provide both maritime and coastal forecasts for polar lows by employing numerical modeling. State-of-the-art, high-resolution, convection permitting simulations are now able to successfully simulate some polar lows (for an example, see Figure 7), and these simulations exhibit good agreement with observation (Sergeev et al., 2017; Wagner et al., 2011). Yet, accurately forecasting of polar lows is still a challenge, and poor forecast skills can be attributed to the relative rapid development and small size of polar lows, lack of observations and radar coverage, and the lack of guidance of the numerical modeling due to limited data-assimilation in these data-sparse regions. The quality of regional forecasting of polar lows depends strongly on size and location of the regional domain selection (Kristiansen, Sorland, Iversen, Borge, & Koltzow, 2011), as well as the applied physical parameterizations and the horizontal resolution (McInnes et al., 2011; Wu & Petty, 2010). The representation of polar lows in operational regional modeling is significantly improved when data-assimilation of satellite products are included (Randriamampianina, Iversen, & Storto, 2011), or by including targeted observations (Irvine, Gray, & Methven, 2011). Furthermore, high-resolution ensembles can significantly improve the quality of the forecasts, with estimates of useful forecasts for polar lows up to 2 days (Aspelien, Iversen, Bremsnes, & Frogner, 2011; Kristiansen et al., 2011). Another challenge for polar low forecasting includes air-sea-ice interactions, as polar lows typically develop downstream of ice-covered areas. The transition zone between ice-covered regions and the open ocean, the marginal ice zone (MIZ), is often poorly represented in numerical models. Yet, the effect of surface fluxes induced over the MIZ can be significant downstream; thus, the representation of inhomogeneities in the MIZ impacts the representation of polar lows in numerical models (Pagowski & Moore, 2001).
Polar low research is characterized by a vast array of case-studies, and it is evident that there are several dynamical pathways for polar low development. Various conceptual models exist to describe polar low development, however verification of these conceptual models and consensus on the underlying dynamical pathways have been restrained due to the lack of observations and suitable datasets to evaluate these conceptual models. Recent developments, including the increased spatial and temporal resolution of reanalysis datasets, such as ERA5 and ASR, and increased remote sensing at high latitudes, such as CloudSat, provide more detailed data to identify common features and derive and test new and existing conceptual models, allowing significant improvements in our dynamical understanding of polar lows. Furthermore, advances in coupled atmosphere-ice-ocean numerical modeling and the development of parameterizations specific for high-latitude environments has the potential to improve our forecasting capacities for polar lows. Recognition of the interaction between mesoscale atmospheric processes, such as polar lows, and the larger climate system, has recently renewed the interest and importance of understanding mesoscale features at high latitudes, and high-resolution, coupled climate models provide tools to evaluate these interactions, including projected changes in future climate scenarios.
- Rasmussen, E. A., & Turner, J. (2003). Polar lows: Mesoscale weather systems in the polar regions. New York, NY: Cambridge University Press.
- Renfrew, I. (2015). Polar lows. In G. R. North, J. Pyle, & F. Zhang (Eds.), Encyclopedia of atmospheric sciences (2nd ed., Vol. 5, pp. 379–386). Amsterdam, The Netherlands: Elsevier.
- Asai, T. (1988). Mesoscale features of heavy snowfalls in Japan Sea coastal regions of Japan (in Japanese). Tenki, 35, 156–161.
- Aspelien, T., Iversen, T., Bremsnes, J. B., & Frogner, I. L. (2011). Short-range probabilistic forecasts from the Norwegian limited-area EPS: Long-term validation and a polar low study. Tellus, 63(3), 564–584.
- Blechschmidt, A., (2008). A 2-year climatology of polar low events over the Nordic Seas from satellite remote sensing. Geophysical Research Letters, 35(9), 815.
- Bobylev, L. P., Zablotski, E. V., Mitnik, L. M., & Mitnik, M. L. (2011). Arctic polar low detection and monitoring using atmospheric water vapor retrievals from satellite passive microwave data. IEEE Transactions on Geoscience and Remote Sensing, 49(9), 3302–3310.
- Bond, N. A., & Shapiro, M. (1991). Polar lows over the Gulf of Alaska in conditions of reverse shear. Monthly Weather Review, 119(2), 551–572.
- Bracegirdle, T. J., & Gray, S. L. (2008). An objective climatology of the dynamical forcing of polar lows in the Nordic seas. International Journal of Climatology, 28(14), 1903–1919.
- Bracegirdle, T. J., & Gray, S. L. (2009). The dynamics of a polar low assessed using potential vorticity inversion. Quarterly Journal of the Royal Meteorological Society, 135(641), 880–893.
- Bresch, J. F., Reed, R. J., & Albright, M. D. (1997). A polar-low development over the Bering Sea: Analysis, numerical simulation, and sensitivity experiments. Monthly Weather Review, 125(12), 3109–3130.
- Bromwich, D. H. (1991). Mesoscale cyclogenesis over the Southwestern Ross Sea linked to strong katabatic winds. Monthly Weather Review, 119(7), 1736–1753.
- Brümmer, B., Müller, G., & Noer, G. (2009). A polar low pair over the Norwegian sea. Monthly Weather Review, 137(8), 2559–2575.
- Businger, S. (1985). The synoptic climatology of polar low outbreaks. Tellus, 37A, 419–432.
- Businger, S. (1987). The synoptic climatology of polar low outbreaks over the Gulf of Alaska and the Bering Sea. Tellus A: Dynamic Meteorology and Oceanography, 39(4), 307–325.
- Businger, S., & Reed, R. J. (1989). Cyclogenesis in cold air masses. Weather and Forecasting, 4(2), 133–156.
- Carleton, A. M. (1985). Satellite climatological aspects of the “polar low” and “instant occlusion.” Tellus A: Dynamic Meteorology and Oceanography, 37(5), 433–450.
- Carleton, A. M., & Song, Y. (1997). Synoptic climatology, and intrahemispheric associations, of cold air mesocyclones in the Australasian sector. Journal of Geophysical Research, 102(D12), 13873–13887.
- Carrasco, J. F., & Bromwich, D. H. (1993). Mesoscale cyclogenesis dynamics over the southwestern Ross Sea, Antarctica. Journal of Geophysical Research, 98(D7), 12973–12995.
- Charney, J. G. (1947). The dynamics of long waves in a baroclinic westerly current. Journal of Meteorogy, 4(5), 135–162.
- Chen, F., & von Storch, H. (2013). Trends and variability of North Pacific polar lows. Advances in Meteorology, 2013, 11.
- Claud, C., Duchiron, B., & Terray, P. (2007). Associations between large-scale atmospheric circulation and polar low developments over the North Atlantic during winter. Journal of Geophysical Research, 112(D12).
- Claud, C., Funatsu, B. M., Noer, G., & Chaboureau, J.-P. (2009). Observation of polar lows by the Advanced Microwave Sounding Unit: potential and limitations. Tellus A: Dynamic Meteorology and Oceanography, 61(2), 264–277.
- Claud, C., Heinemann, G., Raustein, E., & McMurdie, L. (2004). Polar low “le Cygne”: Satellite observations and numerical simulations. Quarterly Journal of the Royal Meteorological Society, 130(598), 1075–1102.
- Condron, A., Bigg, G. R., & Renfrew, I. (2006). Polar mesoscale cyclones in the Northeast Atlantic: Comparing climatologies from ERA-40 and satellite imagery. Monthly Weather Review, 134(5), 1518–1533.
- Condron, A., & Renfrew, I. (2013). The impact of polar mesoscale storms on northeast Atlantic ocean circulation. Nature Geoscience, 6(1), 34–37.
- Craig, G. C., & Cho, H. R. (1988). Cumulus heating and CISK in the extratropical atmosphere. Part I: Polar lows and comma clouds. Journal of Atmospheric Sciences, 45(19), 2622–2640.
- Craig, G. C. (1995). Radiation and polar lows. Quarterly Journal of the Royal Meteorological Society, 121(521), 79–94.
- Deveson, A. C. L., Browning, K. A., & Hewson, T. D. (2002). A classification of FASTEX cyclones using height-attributable quasi-geostrophic vertical motion diagnostic. Quarterly Journal of the Royal Meteorological Society, 128(579), 93–117.
- Douglas, M. W., Shapiro, M., Fedor, L., & Saukkonen, L. (1995). Research aircraft observations of a polar low at the East Greenland ice edge. Monthly Weather Review, 123(1), 5–15.
- Duncan, C. N. (1977). A numerical investigation of polar lows. Quarterly Journal of the Royal Meteorological Society, 103(436), 225–267.
- Duncan, C. N. (1978). Baroclinic instability in a reversed shear flow. Meteorological Magazine, 107(1266), 17–23.
- Eady, E. T. (1949). Long waves and cyclone waves. Tellus, 1(3), 33–52.
- Ese, T., Kanestrom, I., & Pedersen, K. (1988). Climatology of polar lows over the Norwegian and Barents Seas. Tellus A, 40(3), 248–255.
- Emanuel, K. A., & Rotunno, R. (1989). Polar lows as Arctic hurricanes. Tellus A, 41(1), 1–17.
- Fantini, M., & Buzzi, A. (1993). Numerical experiments on a possible mechanism of cyclogenesis in the Antarctic region. Tellus A, 45(2), 99–113.
- Forbes, G. S., & Lottes, W. D. (1985). Classification of mesoscale vortices in polar airstreams and the influence of the large-scale environment on their evolutions. Tellus A, 37(2), 132–155.
- Føre, I., Kristjansson, J. E., Saetre, Ø., Breivik, Ø., Røsting, B., & Shapiro, M. (2011). The full life cycle of a polar low over the Norwegian sea observed by three research aircraft flights. Quarterly Journal of the Royal Meteorological Society, 137(660), 1659–1673.
- Forsythe, J. M., & Haynes, J. M. (2015). CloudSAT observes a Labrador Sea polar low. Bulletin of the American Meteorological Society, 96(8), 1229–1231.
- Gray, S. L., & Craig, G. C., (1998). A simple theoretical model for the intensification of tropical cyclones and polar lows. Quarterly Journal of the Royal Meteorological Society, 124(547), 919–947.
- Grønås, S., Foss, A., & Lystad, M. (1987). Numerical simulations of polar lows in the Norwegian Sea. Tellus A, 39(4), 334–353.
- Grønås, S., & Kvamstø, N. G. (1995). Numerical simulations of the synoptic conditions and development of Arctic outbreak polar lows. Tellus A, 47(5), 797–814.
- Gurvich, I. A. (2013). Intense mesoscale cyclones over the Far Eastern Seas in the cold half of the year according to satellite remote sensing (in Russian) (Doctoral dissertation). V. I. Il’ichev Pacific Oceanolocical Institute, Far Eastern Branch Russian Academy of Sciences.
- Hartmann, J., Kottmeier, C., & Raasch, S. (1997). Roll vortices and boundary-layer development during a cold air outbreak. Boundary-Layer Meteorology, 84(1), 45–65.
- Heinemann, G. (1990). Mesoscale vortices in the Weddell Sea Region (Antarctica). Monthly Weather Review, 118(3), 779–793.
- Heinemann, G. (1996). On the development of wintertime mesoscale cyclones near the sea-ice front in the Arctic and Antarctic. Global Atmosphere and Ocean System, 4, 89–123.
- Hoskins, B., McIntyre, M., & Robertson, A. (1985). On the use and significance of isentropic potential vorticity maps. Quarterly Journal of the Royal Meteorological Society, 111(470), 877–946.
- Irvine, E. A., Gray, S. L., & Methven, J. (2011). Targeted observations of a polar low in the Norwegian Sea. Quarterly Journal of the Royal Meteorological Society, 137(660), 1688–1699.
- Irving, D., Simmonds, I., & Keay, K. (2011). Mesoscale cyclone activity over the ice-free Southern Ocean: 1999–2008. Journal of Climate, 23(20), 5404–5420.
- Joly, A., & Thorpe, A. J. (1989). Warm and occluded fronts in two-dimensional moist baroclinic instability. Quarterly Journal of the Royal Meteorological Society, 115(487), 513–534.
- Klein, T., & Heinemann, G. (2002). Interaction of katabatic winds and mesocyclones near the eastern coast of Greenland. Meteorological Applications, 9(4), 407–422.
- Kolstad, E. W. (2006). A new climatology of favourable conditions for reverse-shear polar lows. Tellus A, 58(3), 344–354.
- Kolstad, E. W. 2011). A global climatology of favourable conditions for polar lows. Quarterly Journal of the Royal Meteorological Society, 137(660), 1749–1761.
- Kristiansen, J., Sorland, S. L., Iversen, T., Borge, D., & Koltzow, M. O. (2011). High-resolution ensemble prediction of a polar low development. Tellus A: Dynamic Meteorology and Oceanography, 63(3), 585–604.
- Landgren, O. A., Batrak, Y, Haugen, J. E.,Stoylen, E., & Iversen, T. (2019). Polar low variability and future projections for the Nordic and Barents Seas. Quarterly Journal of the Royal Meteorological Society, 1–13
- Laffineur, T., Claud, C., Chaboureau, J. P., & Noer, G. (2014). Polar lows over the Nordic Seas: improved representation in ERA-Interim compared to ERA-40 and the impact on downscaled simulations. Monthly Weather Review, 142(6), 2271–2289.
- Linders, T., & Seatra, O. (2010). Can CAPE maintain polar lows? Journal of the Atmospheric Sciences, 67(8), 2559–2571.
- Mailhot, J., Hanley, D., Bilodeau, B., & Hertzman, O. (1996). A numerical case study of a polar low in the Labrador Sea. Tellus A: Dynamic Meteorology and Oceanography, 48(3), 383–402.
- Mallet, P.-E., Claud, C., Cassou, C., Noer, G., & Kodera, K. (2013). Polar lows over the Nordic and Labrador Seas: Synoptic circulation patterns and associations with North Atlantic-Europe wintertime weather regimes. Journal of Geophysical Research, 118(6), 2455–2472.
- Mansfield, D. (1974). Polar lows: The development of baroclinic disturbances in cold air outbreaks. Quarterly Journal of the Royal Meteorological Society, 100(426), 541–554.
- Martin, R., & Moore, G. W. K. (2006). Transition of a synoptic system to a polar low via interaction with the orography of Greenland. Tellus A: Dynamic Meteorology and Oceanography, 58(2), 236–253.
- McInnes, H., Kristiansen, J., Kristjansson, J. E., & Schyberg, H. (2011). The role of horizontal resolution for polar low simulations. Quarterly Journal of the Royal Meteorological Society, 137(660), 1674–1687.
- Melsheimer, C., Frost, T., & Heygster, G. (2016). Detectability of polar mesocyclones and polar lows in data from spaceborne microwave humidity sounders. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 9(1), 326–335.
- Michel, C., Terpstra, A., & Spengler, T. (2018). Polar mesoscale cyclone climatology for the Nordic seas based on era-interim. Journal of Climate, 31(6), 2511–2532.
- Mingalev, I., Orlov, K., & Mingalev, V. (2014). A modeling study of the initial formation of polar lows in the vicinity of the Arctic front. Advances in Meteorology, 2014, 10.
- Mokhov, I. I., Akperov, M. G., Lagun, V. E., & Lutsenko, E. I. (2007). Intense arctic mesocyclones. Atmospheric and Oceanic Physics, 43(3), 259–265.
- Montgomery, M. T., & Farrell, B. F. (1992). Polar low dynamics. Journal of Atmospheric Sciences, 49(24), 2484–2505.
- Moor, R. W., & Vonder Haar, T. H. (2003). Diagnosis of a polar low warm core utilizing the advanced microwave sounding unit. Weather and Forecast, 18(5), 700–711.
- Moore, G. W. K., & Vachon, P. W. (2002). A polar low over the Labrador Sea: Interactions with topography and an upper-level potential vorticity anomaly, and an observation by RADARSAT-1 SAR. Geophysical Research Letters, 29(16), 1773.
- Mullen, S. L. (1979). An investigation of small synoptic-scale cyclones in polar air streams. Monthly Weather Review, 107(12), 1636–1647.
- Murakami, M. (2019). Inner structures of snow clouds over the Sea of Japan observed by instrumented aircraft: A Review. Journal of the Meteorological Society of Japan, 97, 5–38.
- Nagata, M. (1991). Further numerical study on the formation of the convergent cloud band over the Japan Sea in winter. Journal of the Meteorological Society of Japan, 69(3), 419–428.
- Nagata, M. (1993). Meso-β-scale vortices developing along the Japan-Sea Polar-Air Mass Convergence Zone (JPCZ) cloud band: Numerical simulation. Journal of the Meteorological Society of Japan, 71(1), 43–57.
- Noer, G., Saetre, Ø., Lien, T., & Gusdal, Y. (2011). A climatological study of polar lows in the Nordic Seas. Quarterly Journal of the Royal Meteorological Society, 137(660), 1762–1772.
- Nordeng, T. E. (1987). The effect of vertical and slantwise convection on the simulation of polar lows. Tellus A, 39(4), 354–375.
- Nordeng, T. E., & Rasmussen, E. A. (1992). A most beautiful polar low. A case study of a polar low development in the Bear Island region. Tellus A, 44(2), 81–99.
- Nordeng, T. E., & Røsting, B. (2011). A polar low named Vera: the use of potential vorticity diagnostics to assess its development. Quarterly Journal of the Royal Meteorological Society, 137(660), 1790–1803.
- Økland, H. (1986). Heating by organized convection as a source of polar low intensification. Tellus A, 39(4), 397–407.
- Pagowski, M., & Moore, G. W. K. (2001). A numerical study of an extreme cold air outbreak over the Labrador Sea: Sea-ice, air-sea interaction and development of polar lows. Monthly Weather Review, 129, 47–72.
- Papritz, L., & Pfahl, S. (2016). Importance of latent heating in mesocyclones for the decay of cold air outbreaks: A numerical process study from the Pacific sector of the Southern Ocean. Monthly Weather Review, 144(1), 315–336.
- Papritz, L., & Spengler, T. (2017). A Lagrangian climatology of wintertime cold air outbreaks in the Irminger and Nordic Seas and their role in shaping air-sea heat fluxes. Journal of Climate, 30(8), 2717–2737.
- Parker, D. J., & Thorpe, A. J. (1995). Conditional convective heating in a baroclinic atmosphere: A model of convective frontogenesis. Journal of Atmospheric Sciences, 52(10), 1699–1711.
- Pezza, A., Sadler, K., Uotila, P., Vihma, T., Mesquita, M. D. S., & Reid, P. (2015). Southern hemisphere strong polar mesoscale cyclones in high-resolution datasets. Climate Dynamics, 47(5/6), 1647–1660.
- Randriamampianina, R., Iversen, T., & Storto, A. (2011). Exploring the assimilation of IASI radiances in forecasting polar lows. Quarterly Journal of the Royal Meteorological Society, 137(660), 1700–1715.
- Rasmussen, E., (1979). The polar low as an extratropical CISK disturbance. Quarterly Journal of the Royal Meteorological Society, 105(445), 531–549.
- Rasmussen, E. (1985). A case study of a polar low development over the Barents Sea. Tellus A, 37(5), 407–418.
- Rasmussen, E. A., & Turner, J. (2003). Polar lows: Mesoscale weather systems in the polar regions. New York, NY: Cambridge University Press.
- Reed, R. J., & Duncan, C. N. (1987). Baroclinic instability as a mechanism for the serial development of polar lows: A case study. Tellus A, 39(4), 376–384.
- Romero, R., & Emanuel, K. (2017). Climate change and hurricane-like extratropical cyclones: Projections for North Atlantic polar lows and medicanes based on CMIP5 models. Journal of Climate, 30(1), 279–299.
- Sætra, Ø., Gusdal, Y., Eastwood, S., Debernard, J., Isachsen, P.-E., Schyberg, H. . . . Noer, G. (2010). STARS deliverable document D-3, Scientific Analysis Plan. Technical Report. Norwegian Meteorological Institute, Norway.
- Sardie, J. M., & Warner, T. T. (1983). On the mechanism for the development of polar lows. Journal of Atmospheric Sciences, 40(4), 869–881.
- Schar, C., & Davies, H. C. (1990). An instability of mature cold fronts. Journal of Atmospheric Sciences, 47(8), 929–950.
- Sergeev, D., Renfrew, I., Spengler, T., & Dorling, S. (2017). Structure of a shear-line polar low. Quarterly Journal of the Royal Meteorological Society, 143(702), 12–26.
- Sergeev, D., Renfrew, I., & Spengler, T. (2018). Modification of polar low development by orography and sea ice. Monthly Weather Review, 146(10), 3325–3341.
- Shapiro, M., Fedor, L., & Hampel, T. (1987). Research aircraft observations of a polar low over the Norwegian Sea. Tellus A: Dynamic Meteorology and Oceanography, 39(4), 272–306.
- Shapiro, M. A., & Fedor, L. S. (1989). A case study of an ice-edge boundary layer front and polar low development over the Norwegian and Barents Seas. In P. F. Twitchell, E. A. Rasmussen, & K. L. Davidson (Eds.), Polar and artic lows (pp. 257–277). Hampton, VA: Depak.
- Shimada, U., Wada, A., Yamazaki, K., & Kitabatake, N. (2014). Roles of an upper-level cold vortex and low-level baroclinicity in the development of polar lows over the Sea of Japan. Tellus A: Dynamic Meteorology and Oceanography, 66(1).
- Snyder, C., & Lindzen, R. S. (1991). Quasi-geostrophic wave-CISK in an unbounded baroclinic shear. Journal of Atmospheric Sciences, 48(1), 76–86.
- Smirnova, J., & Golubkin, P. (2017). Comparing polar lows in atmospheric reanalysis: ASR vs ERA-Interim. Monthly Weather Review, 145(6), 2375–2383.
- Song, G., Subrahmanyam, M. V., Guo, B., & Panda, J. (2015). A case study of polar low detection using ERS-2 wave mode image. Open Oceanography Journal, 8(1), 28–32.
- Stoll, P. J., Graversen, R.G., Noer, G., & Hodges, K. (2018). An objective global climatology of polar lows based on reanalysis data. Quarterly Journal of the Royal Meteorological Society, 144 (716), 2099–2117.
- Terpstra, A., Michel, C., & Spengler, T. (2016). Forward and reverse shear environments during polar low genesis over the Northeast Atlantic. Monthly Weather Review, 144(4), 1341–1354.
- Terpstra, A., Spengler, T., & Moore, R. W. (2015). Idealised simulations of polar low development in an Arctic moist baroclinic environment. Quarterly Journal of the Royal Meteorological Society, 14(691), 1987–1996.
- Thorncroft, C., & Hoskins, B. (1990). Frontal cyclogenesis. Journal of the Atmospheric Sciences, 47(19), 2317–2336.
- Verezemskaya, P., Tilinaina, N., Gulev, S., Renfrew, I. A., & Lazzara, M. (2017). Southern Ocean mesocyclones and polar lows from manually tracked satellite mosaics. Geophysical Research Letters, 44(15), 7985–7993.
- Wagner, J., Gohm, A., Dörnbrack, A., & Schäfler, A. (2011). The mesoscale structure of a polar low: Airborne lidar measurements and simulations. Quarterly Journal of the Royal Meteorological Society, 137(659), 1516–1531.
- Watanabe, S. I., & Niino, H. (2014). Genesis and development mechanisms of a polar mesocyclone over the Japan Sea. Monthly Weather Review, 142(6), 2248–2270.
- Watanabe, S. I., Niino, H., & Yanase, W. (2016). Climatology of polar mesocyclones over the Sea of Japan using a new objective tracking method. Monthly Weather Review, 144(7), 2503–2515.
- Watanabe, S. I., Niino, H., & Yanase, W. (2017). Structure and environment of polar mesocyclones over the northeastern part of the Japan Sea. Monthly Weather Review, 145(6), 2217–2233.
- Watanabe, S.I., Niino, H., & Yanase, W. (2018). Composite analysis of polar mesocyclones over the western part of the Sea of Japan. Monthly Weather Review, 146(4), 985–1004.
- Wilhelmsen, K., (1985). Climatological study of gale-producing polar lows near Norway. Tellus A: Dynamic Meteorology and Oceanography, 37(5), 451–459.
- Wu, L., & Petty, G. W. (2010). Intercomparison of bulk microphysics schemes in model simulations of polar lows. Monthly Weather Review, 138(6), 2211–2228.
- Wu, L., Martin, J., & G. Petty, (2011). Piecewise potential vorticity diagnosis of the development of a polar low over the Sea of Japan. Tellus A: Dynamic Meteorology and Oceanography, 63(2), 198–211.
- Yanase, W., & Niino, H. (2007). Dependence of polar low development on baroclinicity and physical processes: An idealized high-resolution numerical experiment. Journal of Atmospheric Sciences, 64(9), 3044–3067.
- Yanase, W., Niino, H., & Watanabe, S. (2016). Climatology of polar lows over the Sea of Japan using the JRA-55 reanalysis. Journal of Climate, 29(2), 419–437.
- Zahn, M., & von Storch, H. (2008). A long-term climatology of North Atlantic polar lows. Geophysical Research Letters, 35(22).
- Zahn, M., & von Storch, H. (2010). Decreased frequency of North Atlantic polar lows associated with future climate warming. Nature, 467(7313), 309–312.
- Zappa, G., & Shaffrey, L. (2014). Can polar lows be objectively identified and tracked in the ECMWF operational analysis and the ERA-interim reanalysis? Monthly Weather Review, 142(8), 2596–2608.