1. Introduction
Tropopause polar vortices (TPVs) are a commonly observed feature of the Arctic that exist on daily to monthly time scales. Although both cyclones and anticyclones are observed, we focus here on cyclonic TPVs because these features may excite surface cyclones and, through their cloud fields, affect the Arctic radiation budget. Although numerous and long lived, little is known with regard to factors that control their intensity. These upper-level disturbances are defined by closed material contours on the dynamic tropopause [a potential vorticity (PV) surface], implying that under adiabatic conditions, their fluid properties are isolated from the background environment for which they exist. It is well understood that upper-level disturbances can play a major role in extratropical cyclogenesis (e.g., Holton 2004, chapters 8–9), and that tropopause vortices function in this capacity when captured by the jet stream (e.g., Hakim et al. 1995; Bosart et al. 1996; Hakim et al. 1996). Although many processes may promote the intensification of TPVs, such as Rossby wave breaking, gravity wave breaking, downstream energy propagation, turbulent diffusion, and diabatic effects, lateral wind shear is relatively weak in the high-latitude regions where TPVs are most common. This fact, combined with observations of nearly moist adiabatic lapse rates and clouds in the vicinity of vortex cores (Cavallo and Hakim 2009, hereafter CH09), are indications that diabatic effects may be essential to their maintenance and intensification. In this paper, we seek to quantify the significance of these diabatic effects, and to determine the process with the greatest impact on vortex intensity.
The motivation to understand TPVs derives in part from the observation that surface cyclogenesis precursors may be vortical, rather than wavelike (e.g., Hakim 2000). Such features are well identified by the PV field and, assuming that PV and potential temperature are conserved, a single PV surface that defines the “dynamic tropopause” allows these features to be tracked at one level (e.g., Morgan and Nielsen-Gammon 1998). When viewed from this perspective, upper-tropospheric disturbances can sometimes be tracked for weeks before the occurrence of surface cyclogenesis (e.g., Hakim et al. 1995; Bosart et al. 1996; Hakim et al. 1996). Observations show that these disturbances are characterized by sometimes narrow, downward intrusions of stratospheric air to much lower atmospheric levels (e.g., Uccellini et al. 1985; Davis and Emanuel 1991; Hakim 2000). Hakim (2000) analyzed a 33-yr climatology of these disturbances on the tropopause, and found many were coherent vortices with radii of approximately 500–800 km, tropopause potential temperature amplitudes of at least 8 K, and a downward intrusion of the tropopause to approximately 600 hPa. Using a 51-yr climatology (1948–99), Hakim and Canavan (2005) subsequently found that TPVs exhibited preferential track locations, moved equatorward on average, and could have lifetimes of over 1 month.
The climatology of Hakim and Canavan (2005) was extended by CH09, where it was found that TPV cyclogenesis, as well as cyclone intensification, occur in preferential locations, with the highest probability over Baffin Island in the Canadian Arctic. In addition, no significant relationship with the large-scale flow patterns was identified, and observations from radiosondes near the vortex cores suggested the importance of diabatic effects, as evidenced by the prevalence of moist adiabatic temperature profiles through the depth of the troposphere. CH09 performed a complete PV budget analysis of a single case of an intensifying cyclonic TPV and showed that radiative cooling, enhanced near the top of the cloud layer, acted to intensify the vortex while latent heating promoted weakening. Cloud-top radiative cooling dominated, leading to a net intensification of the vortex. The goal of the present study is to assess the generality of these results by testing the hypothesis that radiative processes are responsible for the maintenance and intensification of cyclonic TPVs. A PV diagnostic approach is applied to numerical simulations of a large sample of cases to gather the composite diabatic tendencies associated with TPVs in a region of the Arctic where they occur most frequently.
The remainder of the paper is organized as follows. The methods used to quantify TPV intensification using a PV diagnostics approach are discussed in section 2. A description of the numerical model and the procedure for identifying TPVs in the model data are also provided in section 2. Verification of the model simulations is given in section 3. Results from a large sample of TPVs are examined and discussed in section 4, including a component-wise examination of the diabatic PV tendencies. The summary and conclusions are given in section 5.
2. Method
a. Model setup
The composite structure of cyclonic TPVs is evaluated during the time period from 2 August 2007 to 31 July 2009 for a limited-area domain over a region of the Arctic where the climatological frequency of cyclonic TPVs is high. We use the Advanced Research Weather (ARW) version of the Weather Research and Forecasting (WRF) model version 2.2.1 (Skamarock et al. 2005) and the WRF Preprocessing System (WPS) version 2.2. Time integration is achieved by a third-order accurate Runge–Kutta method, and horizontal advection by a fifth-order accurate scheme. A horizontal grid spacing of Δx = Δy = 30 km is used, with 31 vertical levels and a model time step of Δt = 120 s. The following physics schemes are used: Rapid Radiative Transfer Model (RRTM) longwave radiation (Mlawer et al. 1997), the National Aeronautics and Space Administration (NASA) Goddard shortwave radiation (Chou and Suarez 1994), WRF single-moment (WSM) 5-class microphysics (Hong et al. 2004), Kain–Fritsch cumulus convection (Kain and Fritsch 1993), Yonsei University (YSU) planetary boundary layer (Hong and Pan 1996), thermal diffusion surface physics (Chen and Dudhia 2001a,b), and Monin–Obukhov surface layer physics (Paulson 1970; Dyer and Hicks 1970; Webb 1970). Forecasts are initialized using Global Forecasting System (GFS) analyses at 0000 UTC daily, with boundary conditions derived from GFS forecasts every 3 h. The simulations are performed on a polar stereographic map projection, with a center latitude and longitude of 68.46°N and 81.14°W with 120 x and y grid points, respectively.
The model domain is shown in Fig. 1, which corresponds to a region in the Canadian Arctic. Regions of high density for cyclonic TPV intensity change have previously been determined by CH09 using 2.5° National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis data every 6 h during 1948–99 (Kalnay et al. 1996). Vortices were tracked using a tracking algorithm described in Hakim and Canavan (2005), and intensification refers to the location of the greatest increase in amplitude during a 24-h period along a unique vortex track. The data show that the densest region is found near northern Baffin Island in the Canadian Arctic. To maximize the cyclone sample potential, a domain was chosen to surround this most likely climatological region. Consideration of the domain location is also based on the expected southward motion of cyclones out of the Arctic shown by Hakim and Canavan (2005). Another advantage of this particular domain is the presence a relatively dense radiosonde network, which provides an additional observation source in the model analyses used for verification. The model simulations were performed in real time during the 2-yr period.
b. Vortex identification
Cyclonic vortices are objectively identified using the following procedure. First, the minimum potential temperature on the dynamic tropopause, defined as the 2-PVU (1 PVU ≡ 1 × 10−6 K kg−1 s−1 m2) surface, is determined at the analysis time. If the minimum value is at least two standard deviations below the domain mean, then the data are further considered as a possible vortex to be included in the composite sample. This threshold is intended to retain only those cyclones with minimum tropopause potential temperature values significantly lower than the background; however, if the variance about the domain-mean value is small, the vortex amplitude does not need to be large to meet this condition. A vortex is included in the composite sample if it lies entirely within the bounds of the domain. This condition is satisfied if a sign reversal in the gradient of potential temperature can be found in each direction along eight equally spaced radials. The minimum tropopause potential temperature value along the eight radials at the location of the sign reversal is defined as the last closed contour, and vortex amplitude is defined as the absolute difference between the last closed contour and the core value. Composite maps are created by averaging all fields along a west–east line relative to the grid point corresponding to the vortex core. No data within five grid points of the domain boundaries are used at any time and are not included in the composite sample. Since the fields are averaged relative to the vortex core on a fixed domain, the distance of the vortices from the domain edge is different for each sample, and the number of data points is greatest within the composited vortex and decreases radially outward. Because of the limited geographical area covered by the domain, it is not possible to objectively track cyclonic TPVs through their entire life span. Although vortices can be tracked within the domain, we choose instead to focus on their composite structure with attention given to the mechanisms of intensity change. Therefore, the number of vortex samples does not correspond to the number of individual vortex tracks, and simply represents a count of the vortex samples using the criteria above. With these criteria, a total of 568 vortex samples have been obtained during the period 2 August 2007–31 July 2009, out of 683 total model simulations.
To assess the impact of the filtering process described previously, which is designed to isolate well-resolved vortices, an unfiltered set of vortices is defined using the same criteria described above, except that all vortices in the domain are included and no amplitude threshold is applied. Note that in the unfiltered set, multiple vortices can be associated with a single analysis, whereas only one vortex is retained using the filtering criteria. Results show that cyclones have an average radius of 465 km, while the unfiltered cyclones have an average radius of 358 km taken from 1997 samples (Fig. 2a). Radii from both samples are generally less than 1000 km with the peak in the distribution occurring near the mean. Small cyclones are preferentially eliminated by the filter, with the peak reduction at a radius of 275 km. The average cyclone amplitude is 20.7 K in the composite data and 14.3 K in the unfiltered sample (Fig. 2b). These results indicate that filtering eliminates mainly smaller-amplitude, smaller-scale vortices, with the greatest reduction in vortices with amplitudes around 8 K; some of these features are grid-scale noise, whereas others may be physically relevant. In both samples, a rather long tail is present in the amplitude distribution, with cyclone amplitudes reaching 50 K. We now proceed to assess the bias in the model simulations by comparison to GFS analyses.
3. Tropopause polar cyclone verification
WRF tropopause potential temperature forecast biases are shown in Fig. 3. Biases in both the 24- and 48-h WRF forecasts are generally less than 1.5 K in magnitude and there is no clear sign preference in the domain. Biases are higher in localized areas, and appear to be largest in magnitude near radiosonde stations. This result probably reflects the assimilation process, since the GFS analyses are based upon a weighting of observations and a 6-h forecast (Derber et al. 1991), whereas no observation adjustments are made in the WRF forecasts. For 48-h forecasts, the bias pattern is roughly similar to, but larger in magnitude than the 24-h forecasts. The largest biases are present near Cambridge Bay, Nunavut, The Pas, Manitoba, and Aasiaat, Greenland. Given that both tropopause potential temperature and EPV are highly derived quantities, we conclude that WRF forecasts are reasonably consistent, on average, with the GFS analyses.
Next we consider vortex-relative bias, where fields are averaged relative to each vortex core. Both the WRF and GFS vortex cores are found using the same filtering criteria described above, and furthermore must be located within 650 km of each other to be considered the same vortex. Sample frequency decreases away from the vortex core so that there is less confidence in the bias near the domain edges. At the vortex core on the tropopause, 24-h forecasts of WRF potential temperature are nearly 1 K lower than GFS analyses, while biases decrease radially outward, indicating a slightly stronger vortex in the WRF forecasts than is analyzed, on average (Fig. 4a). Note that this apparent difference in vortex strength could arise from interpolation of the relatively course GFS fields to the finer WRF grid. Away from the vortex, WRF forecasts have a positive bias at nearly all levels, with the smallest biases near the tropopause. Note that, especially away from the vortex core, there is an implied bias in lapse rates, with WRF lapse rates slightly greater (less) than GFS in the upper (lower) troposphere. WRF 24-h water vapor mixing ratio is biased to smaller values to the east of the vortex core, and along the tropopause (Fig. 4b). Bias in 500-hPa geopotential height (Fig. 4c) reflects lower 500-hPa heights in the 24-h WRF forecasts, and a southeastward displacement of the location of the 500-hPa height minimum compared to GFS analyses. A similar pattern is apparent in tropopause potential temperature, with values southeast of the vortex core approximately 2 K lower in WRF 24-h forecasts, again reflecting displacement of the vortex core (Fig. 4d). The biases observed here are potentially attributable to some extent to different horizontal grid resolution. The WRF grid is evenly spaced with 30-km horizontal resolution, whereas the GFS grid spacing is 111 km × ∼40 km in the zonal and meridional directions, respectively, at the center of the domain. In any event, the bias appears to be associated mainly with a position shift of the vortex and is small in magnitude compared to the results of the diagnostic analysis presented in the following section.
4. Composite structure
a. Annual structure
Tropopause polar cyclone density during the 2-yr period is greatest along the northwestern tip of Baffin Island, Canada, as well as in the northern portions of Ellesmere Island, Canada, and northern Greenland (Fig. 5). Other areas of relatively large density include northern Baffin Bay and the northern Canadian Archipelago region. The region of large cyclone density over northern Baffin Island is located near the region of maximum cyclone intensification (cf. Fig. 1) and event density (Hakim and Canavan 2005) found in the NCEP–NCAR reanalysis climatology.
To compare the composite vortex data to the background environment, unconditional1 composites are also performed along a fixed cross section, oriented from west to east near the center of the domain, in close proximity to radiosonde stations near Fort Smith, Baker Lake, Coral Harbour, and Iqaluit (see Fig. 3). The fixed cross section is referred to here as the background cross section, and anomalies are defined as the difference between the vortex-relative composite and the background composite (vortex composite − background composite). Although vortices may be sampled as part of the background, we expect their distribution is random, having a minor effect on the background; results shown subsequently confirm this expectation. A fixed point near the center of the cross section, near the Coral Harbour radiosonde station, is chosen to represent a background sounding for comparison with profiles in the vortex core. For the vortex-core composite, the tropopause (2-PVU surface) is located near ∼600 hPa, which compares with ∼350 hPa in the background (Fig. 6a). Temperature anomalies with respect to the background sounding are negative in the troposphere and positive in the stratosphere above ∼375 hPa, with the greatest anomalies located around 700 and 250 hPa, respectively (Fig. 6b). Relative humidity anomalies follow roughly the same pattern as temperature but are of opposite sign, with the greatest anomalies located around 800 and 375 hPa in the troposphere and stratosphere, respectively. A positive EPV anomaly is found at nearly all levels, with a peak near 300 hPa.
Seasonal variations are evident in the vortex core profiles (Fig. 7). In winter and spring, tropospheric lapse rates are relatively small, with a poorly distinguished tropopause near 600 hPa (Figs. 7a,b). Moreover, only a relatively subtle temperature and dewpoint transition exists across the tropopause during winter and spring. Tropospheric lapse rates are larger in summer, with an average in the vortex of ∼7.4 K km−1, with large relative humidity, especially in the lower troposphere (Fig. 7c). The tropopause is well defined in summer, located near 500 hPa. The well-defined summer tropopause is consistent with the findings by Randel et al. (2007), where data from GPS radio occultation observations showed a stronger tropopause temperature inversion in polar regions during the summer. Furthermore, they suggested that the strength of the inversion may be related to the radiative effects of water vapor and ozone, which also exhibit similar seasonal variations. The radiative seasonal variations and its impact on the EPV tendency structure of cyclonic TPVs will be explored in greater detail in section 4b. In autumn, the vortex tropopause lowers to ∼575 hPa, and tropospheric lapse rates decrease to ∼6.6 K km−1 (Fig. 7d). In all seasons, the difference in tropopause pressure between the vortex core and background environment varies little from ∼200 to 250 hPa, with a slightly larger value in the autumn.
Regarding the horizontal structure of the composite vortex, Fig. 8a shows that the tropopause potential temperature is 280 K in the vortex core averaged over all seasons. The larger gradient south of the vortex core results from the background meridional potential temperature gradient, with warmer potential temperature in the southern portion of the domain on average. The 500-hPa geopotential height exhibits a similar pattern, with the lowest height in the vortex core, 5110 m, and the highest height, ∼5350 m, located roughly 1500 km southwest of the vortex core of (Fig. 8b).
Vertical east–west cross sections through the composite vortex reveal cold (warm) temperature anomalies of −8.5 K (+5.5 K) relative to the background are centered on the vortex under (above) the tropopause (Fig. 9a). Anomalous winds are cyclonic around the vortex core, with maximum speeds of about 14 m s−1 400 km from the vortex core near the steepest slope in the tropopause (Fig. 9b). A positive EPV anomaly is located above the composite tropopause, with the greatest anomalies of ∼4 PVU near the background tropopause (Fig. 9c). A downward intrusion of relatively drier, stratospheric, air is evident as a negative relative humidity anomaly of over 25% centered near the background tropopause in the vortex core (Fig. 9d). In contrast, relative humidity is anomalously high under the vortex tropopause, and slopes upward and eastward relative to the vortex. This pattern of anomalously high (low) relative humidity under (over) the vortex core increases the relative humidity gradient across the tropopause (cf. Fig. 6b), which is important for the radiative creation of EPV as will be shown below.
Composite west–east cross sections of the potential temperature diabatic tendency terms on the right-hand side of (3) are shown in Fig. 10. Total diabatic heating rates are negative, except in the lowermost troposphere near the region with anomalously high relative humidity (Fig. 10a). The largest cooling rates are found along the tropopause, particularly inside and to the east of the vortex. This pattern results in anomalous diabatic cooling (heating) centered on the vortex (background) tropopause (Fig. 10b). Radiative cooling rates are less than −0.6 K day−1 at all locations, with the largest cooling rate of −2 K day−1 on the tropopause in the vortex core (Fig. 10c). The anomalous radiative heating rate pattern is similar to the anomalous total diabatic heating rate pattern, with cooling (heating) centered on the vortex (background) tropopause (Fig. 10d). Latent heating rates are positive above ∼925 hPa and greatest east of the vortex core, reaching around +2 K day−1 (Fig. 10e). Below this level, latent cooling is consistent with sublimation or evaporation below a precipitating cloud layer, which are frequently observed in the Arctic throughout the year (e.g., Curry et al. 1996; Walsh et al. 2009). Anomalously large latent heating rates are present below and to the east of the vortex core (Fig. 10f) in the region of anomalously large relative humidity.
The anomalous latent heating pattern suggests an asymmetry in the vertical motion field relative to the vortex core. Indeed, results show ascent is greatest ∼400 km east of the vortex core near 600 hPa, and there is an additional region of ascent ∼800 km east of the vortex core (Fig. 11a). A region of relatively weaker descent is centered ∼300 km west of the vortex core near the tropopause, extending across the vortex core (Fig. 11a). Because of the latitudinal range of the domain, we expect a zonal gradient in temperature to be balanced by vertical wind shear, which we subsequently refer to as the background shear. We hypothesize that this vertical motion pattern is forced both radiatively and by the mean background shear across the vortex. Specifically, the anomalous gradient in radiative heating (see Fig. 10d) is expected to force a nearly symmetric circulation with subsidence in the vortex core surrounded by ascent on the vortex peripheries (e.g., Eliassen 1952; Zierl and Wirth 1997), whereas the presence of mean shear across the vortex is expected to produce a dipole of vertical motion across the vortex with rising motion located downshear of the vortex core. To qualitatively assess the relative size of these contributions, the vertical motion is separated into components symmetric (Fig. 11b) and asymmetric (Fig. 11c) relative to the vortex core. The asymmetric component of the vertical motion is defined at each level by wa(x) = w(x) − w(−x), and the symmetric component by ws(x) = w(x) − wa(x), where w is the vertical motion field and x is the zonal direction with the origin (x = 0) located at the vortex core. The asymmetric portion dominates, and is qualitatively in accord with eastward motion of the vortex, based on expectations of positive (negative) geostrophic vorticity advection by the thermal wind [e.g., Bluestein 1992, his Eq. (5.7.42)] east (west) of the vortex core where air is rising (sinking).
The corresponding composite EPV anomaly cross sections from the right-hand side of (2) are shown in Fig. 12. Anomalous EPV generation of 0.28 PVU day−1 at 450 hPa is located roughly 100 hPa above the vortex tropopause (Fig. 12a). Anomalous EPV generation extends downward below the background tropopause to nearly 600 hPa and upward to approximately 350 hPa. A similar pattern is present in the anomalous EPV tendency from the radiative diabatic component, except with slightly larger magnitude for EPV generation in the vortex compared to the full diabatic term (Fig. 12b). EPV destruction from latent heating extends from ∼750 hPa upward to 400 hPa (Fig. 12c). EPV tendencies from all other diabatic components are small in comparison to the full diabatic term (Fig. 12d).
These results show that, for a large sample of TPVs in the Canadian Arctic, cyclonic EPV generation is present in the vortex, and therefore there is a tendency for cyclonic TPVs to intensify on average due to radiative processes. The anomalous EPV tendency pattern may be understood by appealing to the vertical vorticity component term in (2), so that DΠ/Dt ≃ (ζa/ρ)(∂θ̇/∂z). Comparison of Figs. 12a,b to 10b,d suggest that the EPV generation is largely established by the enhanced vertical heating gradient from anomalous radiative cooling (heating) at the vortex (background) tropopause. As there is also a qualitatively similar relative humidity pattern with an enhanced vertical relative humidity gradient near the tropopause (cf. Fig. 9d), we hypothesize that changes in the vertical moisture distribution play an important role in creating the radiative anomalies that result in the vortex intensification seen here. The specific role that changes in vertical moisture gradients may have on vortex intensity change is beyond the scope of the present study, and will be explored in future research.
b. Seasonal variability
While it is well known that Arctic temperature and radiative fluxes have large seasonal variations (e.g., Peixoto and Oort 1992, see their chapter 6; Serreze et al. 2007), it is unclear if or how these variations affect TPV intensification. For example, with seasonal surface temperature variations of over 50 K in some regions of the Arctic (e.g., Peixoto and Oort 1992, see their chapter 7), large variations in latent heating might be expected due to the exponential relationship between temperature and saturation vapor pressure. Here we briefly examine the seasonal dependence of TPV structure and intensity tendency for the 2-yr period of interest.
Comparing seasonal composite tropopause potential temperature anomalies reveals that they are largest during the winter and spring, with anomalies of −25 and −27 K, respectively (Figs. 13a,b), compared to −21 and −22 K in summer and autumn, respectively (Figs. 13c,d). This result is similar to those of Hakim and Canavan (2005), where it was found that cyclone amplitude is smallest in summer. The horizontal scale of the composite vortex anomalies varies little, but is smallest in winter and autumn (Figs. 13a–d). Comparing vertical cross sections by season indicates that the magnitude of the temperature anomalies is considerably larger in summer, particularly in the lower stratosphere, with anomalies of +6.5 K compared to +2.5 K in winter (Figs. 14a,b). We hypothesize that this additional warming is due to the absorption of solar radiation by ozone, which is absent in the winter. Relative humidity anomalies exhibit a pattern similar to temperature, with much larger anomalies in the summer (Figs. 14c,d). In contrast, radiative heating anomalies above the tropopause are of nearly the same magnitude in both the winter and summer, except the vertical extent is larger in the summer. Moreover, stratospheric warming is found on larger horizontal scales than the vortex anomalies, which again is potentially due to ozone. Radiative cooling is considerably stronger near the vortex tropopause in summer (Figs. 14e,f), with the anomalies shifted upward by ∼100 hPa relative to winter. This shift appears to be collocated with the area of anomalously large vertical gradient of relative humidity near the vortex core (cf. Figs. 14c,d).
The upward shift in the radiative heating anomalies is important because it implies an upward shift in the region of EPV generation into the lower stratosphere in summer. Results largely confirm this expectation, with smaller EPV tendencies roughly two-thirds as large in magnitude in winter, except for slightly larger values on the tropopause inside the vortex (cf. Figs. 15a,b). A similar pattern is found for the EPV tendency due to radiation (cf. Figs. 15c,d), with the EPV tendency roughly half as large in magnitude in winter. The EPV tendency from latent heating also exhibits a strong seasonal difference, with maximum EPV destruction on the tropopause only about 40% of the summer value (Figs. 15e,f). These seasonal differences are consistent with those due to warmer clouds of greater optical thickness, with higher latent heating rates inside the cloud, and stronger cloud–radiative effects in regions of increased hydrometeor gradients as would be expected in summer. These results suggest that an interesting sensitivity may exist in the relationship between radiative and latent heating processes. With globally averaged surface temperatures predicted to continue to increase over the next century (Solomon et al. 2007), and Arctic temperature increases expected to be greater than the global average due at least in part to an ice–albedo positive feedback (e.g., Cess et al. 1991), further examination of this sensitivity is of particular interest for future work.
5. Summary and conclusions
A previous case study showed that TPV intensification was due to radiative processes (CH09), which motivated the hypothesis explored here that cyclonic TPV maintenance and intensification is radiatively driven. Diabatic mechanisms that affect cyclonic tropopause polar vortex (TPV) intensity were quantified using a mesoscale numerical model and a potential vorticity framework on a large sample of cases. The geographical region of interest was chosen based on previous studies that identified regions of largest TPV density (Hakim and Canavan 2005) and intensification (CH09) near Baffin Island in the Canadian Arctic. Objective criteria were used to identify vortices in the model gridded data, yielding a sample of 568 events.
The composite results support the hypothesis, showing that there is an average tendency for TPVs to diabatically intensify in the Canadian Arctic, and that radiation is the dominant component of intensification. While latent heating had a significant affect on vortex intensity by destroying EPV near the tropopause, the magnitude was approximately two-thirds the magnitude of radiation; all other diabatic tendencies were small in comparison. Moreover, seasonal composite EPV tendencies showed that radiation intensifies vortices on average during both the summer and winter, despite the fact that solar radiation is absent (nearly continuous) in winter (summer). Furthermore, while EPV destruction due to latent heating increases in the summer, radiative EPV creation also increases. However, the radiatively generated region of EPV in the vortex shifts slightly upward, away from the tropopause in the summer, and therefore no additional intensification is apparent at that level. Composite diabatic tendencies further revealed that the radiative EPV creation in the vortex core is associated with an enhanced vertical radiative heating gradient across the vortex tropopause. This enhanced gradient arises from anomalously large radiative cooling centered just under the vortex tropopause, and anomalously small radiative cooling in the lower stratosphere near the level of the background tropopause. The radiative cooling pattern in the vortex is qualitatively similar to the composite relative humidity anomalies.
The vertical distribution of the background longwave radiative cooling rates found here agree reasonably well with those predicted by Manabe and Strickler (1964), and the particular longwave radiation scheme (Mlawer et al. 1997) used to obtain the results here has been validated with line-by-line calculations, which have also been validated with observations (Clough and Iacono 1995). With regard to latent heating, the resulting EPV tendencies exhibit the same patterns of EPV destruction (creation) in upper (lower) atmospheric levels seen in midlatitude storm systems (e.g., Davis and Emanuel 1991; Davis 1992; Stoelinga 1996; Lackmann 2002). Although the magnitude of the latent heating rates is considerably smaller here, they are qualitatively consistent with expectations given the relatively colder Arctic temperatures. Low-level latent heating patterns suggest frequently occurring clouds, consistent with observations from radar and lidar, which show a high frequency of cloud cover in the Arctic, with an annual cloud cover frequency of 85%, and a cloud base most frequently 0.25–1.0 km above the ground (Intrieri et al. 2002).
In terms of vortex structure, the Arctic tropopause cyclones studied here have properties comparable to coherent tropopause disturbances in the Northern Hemisphere presented in Hakim (2000). While the coherent tropopause disturbances of Hakim (2000) were defined using different filtering techniques than the Arctic tropopause cyclones shown here, and were not regionally constrained, many of the characteristics are similar, most notably a lowered tropopause to ∼550 hPa, and tropopause potential temperature anomalies around 20 K. Furthermore, cyclonic TPVs exhibit slightly smaller radii scales in the present study, with mean radii of 465 km compared to 500–800 km. Vertical motion exhibits an asymmetric dipole pattern across the vortex, with rising air downshear and sinking air upshear of the vortex. Sinking air overlaps the vortex core, consistent with enhancement of vorticity near the tropopause by vortex stretching. We attribute the dipole pattern to a balanced circulation due to the vertical shear near the vortex associated with the average meridional temperature gradient, and sinking air in the vortex to a balanced circulation in response to radiative cooling near the vortex tropopause.
Results here suggest a possible connection between vortex intensification and radiative processes associated with moisture anomalies, especially with anomalously dry stratospheric air and moist tropospheric air. Latent heating, which adversely affects vortex intensity, is limited by the cold temperatures of the Arctic, and this may be a factor that inhibits the intensification of cold-core vortices in warmer background environments. Future research will be devoted to idealized experiments designed to explore these relationships. Finally, as the Arctic warms and sea ice declines (e.g., Comiso et al. 2008), the relative importance of latent heating may increase, and thus could have an effect on vortex intensification. This is also a natural topic for a future study.
Acknowledgments
This work represents a portion of the first author’s Ph.D. dissertation at the University of Washington. The authors thank Profs. Dale Durran, Qiang Fu, and Mike Wallace for thoughtful discussions and comments regarding this work. This research was sponsored by the National Science Foundation through Award ATM-0552004 made to the University of Washington.
REFERENCES
Bluestein, H. B., 1992: Principles of Kinematics and Dynamics. Vol. I, Synoptic-Dynamic Meteorology in Midlatitudes, Oxford University Press, 431 pp.
Bosart, L. F., G. J. Hakim, K. R. Tyle, M. A. Bedrick, W. E. Bracken, J. J. Dickinson, and D. M. Schultz, 1996: Large-scale antecedent conditions associated with the 12–14 March 1993 cyclone (“Superstorm ’93”) over eastern North America. Mon. Wea. Rev., 124 , 1865–1891.
Cavallo, S. M., and G. J. Hakim, 2009: Potential vorticity diagnosis of a tropopause polar cyclone. Mon. Wea. Rev., 137 , 1358–1371.
Cess, R. D., and Coauthors, 1991: Interpretation of snow-climate feedback as produced by 17 general-circulation models. Science, 253 (5022) 888–892.
Chen, F., and J. Dudhia, 2001a: Coupling and advanced land surface-hydrology model with the Penn State–NCAR MM5 modeling system. Part I: Model implementation and sensitivity. Mon. Wea. Rev., 129 , 569–585.
Chen, F., and J. Dudhia, 2001b: Coupling and advanced land surface-hydrology model with the Penn State–NCAR MM5 modeling system. Part I: Preliminary model validation. Mon. Wea. Rev., 129 , 587–604.
Chou, M-D., and M. J. Suarez, 1994: An efficient thermal infrared radiation parameterization for use in general circulation models. NASA Tech. Memo. 104606, Rep. 3, 85 pp.
Clough, S. A., and M. J. Iacono, 1995: Line-by-line calculation of atmospheric fluxes and cooling rates. 2. Application to carbon dioxide, ozone, methane, nitrous oxide, and the halocarbons. J. Geophys. Res., 100 , (D8). 16519–16535.
Comiso, J. C., C. L. Parkinson, R. Gersten, and L. Stock, 2008: Accelerated decline in the Arctic sea ice cover. Geophys. Res. Lett., 35 , L01703. doi:10.1029/2007GL031972.
Curry, J. A., W. B. Rossow, D. Randall, and J. L. Schramm, 1996: Overview of Arctic cloud and radiation characteristics. J. Climate, 9 , 1731–1764.
Davis, C. A., 1992: A potential-vorticity diagnosis of the importance of initial structure and condensational heating in observed extratropical cyclogenesis. Mon. Wea. Rev., 120 , 2409–2428.
Davis, C. A., and K. A. Emanuel, 1991: Potential vorticity diagnostics of cyclogenesis. Mon. Wea. Rev., 119 , 1929–1953.
Derber, J. C., D. F. Parrish, and S. J. Lord, 1991: The new global operational analysis system at the National Meteorological Center. Wea. Forecasting, 6 , 538–547.
Dyer, A. J., and B. B. Hicks, 1970: Flux-gradient relationships in the constant flux layer. Quart. J. Roy. Meteor. Soc., 96 , 715–721.
Eliassen, A., 1952: Slow thermally or frictionally controlled meridional circulation in a circular vortex. Astrophys. Norv., 5 , 19–60.
Hakim, G. J., 2000: Climatology of coherent structures on the extratropical tropopause. Mon. Wea. Rev., 128 , 385–406.
Hakim, G. J., and A. K. Canavan, 2005: Observed cyclone-anticyclone tropopause asymmetries. J. Atmos. Sci., 62 , 231–240.
Hakim, G. J., D. Keyser, and L. F. Bosart, 1995: The Ohio Valley wave-merger cyclogenesis event of 25–26 January 1978. Part I: Multiscale case study. Mon. Wea. Rev., 123 , 2663–2692.
Hakim, G. J., D. Keyser, and L. F. Bosart, 1996: The Ohio Valley wave-merger cyclogenesis event of 25–26 January 1978. Part II: Diagnosis using quasigeostrophic potential vorticity inversion. Mon. Wea. Rev., 124 , 2176–2205.
Holton, J. R., 2004: An Introduction to Dynamic Meteorology. 4th ed. Elsevier Academic Press, 535 pp.
Hong, S-Y., and H-L. Pan, 1996: Nonlocal boundary layer vertical diffusion in a medium-range forecast model. Mon. Wea. Rev., 124 , 2322–2339.
Hong, S-Y., J. Dudhia, and S-H. Chen, 2004: A revised approach to ice microphysical processes for the bulk parameterization of clouds and precipitation. Mon. Wea. Rev., 132 , 103–120.
Intrieri, J. M., C. W. Fairall, M. D. Shupe, P. O. G. Persson, E. L. Andreas, P. S. Guest, and R. E. Moritz, 2002: An annual cycle of Arctic surface cloud forcing at SHEBA. J. Geophys. Res., 107 , 8039. doi:10.1029/2000JC000423.
Kain, J. S., and J. M. Fritsch, 1993: Convective parameterization for mesoscale models: The Kain–Fritsch scheme. The Representation of Cumulus Convection in Numerical Models, Meteor. Monogr., No. 24, Amer. Meteor. Soc., 165–170.
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77 , 437–471.
Lackmann, G. M., 2002: Cold-frontal potential vorticity maxima, the low-level jet, and moisture transport in extratropical cyclones. Mon. Wea. Rev., 130 , 59–74.
Manabe, S., and R. F. Strickler, 1964: Thermal equilibrium of the atmosphere with a convective adjustment. J. Atmos. Sci., 21 , 361–385.
Mlawer, E. J., S. J. Taubman, P. D. Brown, M. J. Iacono, and S. A. Clough, 1997: Radiative transfer for inhomogeneous atmosphere: RRTM, a validated correlated-k model for the longwave. J. Geophys. Res., 102 , (D14). 16663–16682.
Morgan, M. C., and J. W. Nielsen-Gammon, 1998: Using tropopause maps to diagnose midlatitude weather systems. Mon. Wea. Rev., 126 , 2555–2579.
Paulson, C. A., 1970: The mathematical representation of wind speed and temperature profiles in the unstable atmospheric surface layer. J. Appl. Meteor., 9 , 857–861.
Pedlosky, J., 1998: Geophysical Fluid Dynamics. 2nd ed. Springer, 728 pp.
Peixoto, J. P., and A. H. Oort, 1992: Physics of Climate. 2nd ed. Springer, 520 pp.
Randel, W. J., F. Wu, and P. Forster, 2007: The extratropical inversion layer: Global observations with GPS data, and a radiative forcing mechanism. J. Atmos. Sci., 64 , 4489–4496.
Serreze, M. C., A. P. Barrett, A. G. Slater, M. Steele, J. Zhang, and K. E. Trenberth, 2007: The large-scale energy budget of the Arctic. J. Geophys. Res., 112 , D11122. doi:10.1029/2006JD008230.
Skamarock, W. C., J. B. Klemp, J. Dudhia, D. O. Gill, D. M. Barker, W. Wang, and J. G. Powers, 2005: A description of the advanced research WRF version 2. Tech. Rep., NCAR, 88 pp.
Solomon, S., D. Qin, M. Manning, M. Marquis, K. Averyt, M. M. B. Tignor, H. L. Miller Jr., and Z. Chen, Eds. 2007: Climate Change 2007: The Physical Science Basis. Cambridge University Press, 996 pp.
Stoelinga, M. T., 1996: A potential vorticity–based study of the role of diabatic heating and friction in a numerically simulated baroclinic cyclone. Mon. Wea. Rev., 124 , 849–874.
Uccellini, L. W., D. Keyser, K. F. Brill, and C. H. Wash, 1985: The President’s Day cyclone of 18–19 February 1979: Influence of upstream trough amplification and associated tropopause folding on rapid cyclogenesis. Mon. Wea. Rev., 113 , 962–988.
Walsh, J. E., W. L. Chapman, and D. H. Portis, 2009: Arctic cloud fraction and radiative fluxes in atmospheric reanalyses. J. Climate, 22 , 2316–2334.
Webb, E. K., 1970: Profile relationships: The log-linear range, and extension to strong stability. Quart. J. Roy. Meteor. Soc., 96 , 67–90.
Zierl, B., and V. Wirth, 1997: The influence of radiation on tropopause behaviour and stratosphere-troposphere exchange in an upper tropospheric anticyclone. J. Geophys. Res., 102 , 23883–23894.
Area-weighted occurrence of intensifying tropopause polar cyclones during the period 1948–99 (contours) with the model domain boundary (box). Values are equal to the number of unique vortices within a 5° latitude × 15° longitude box divided by the cosine of latitude (contour interval of 50). Dashed lines correspond to the 50, 100, and 150 contours, while solid lines correspond to the 200, 250, and 300 contours. Only vortices lasting at least 2 days and that spent at least 60% of their lifetimes north of 65°N latitude are included. Locations of radiosonde stations are denoted by the “+” symbols.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3371.1
Tropopause polar cyclonic vortex (a) radius and (b) amplitude during 2 Aug 2007–31 Jul 2009. The filtered cyclone distribution discussed in the text is given by solid lines, and the total vortex distribution is given by the dashed line.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3371.1
The WRF tropopause potential temperature model bias for the (a) 24- and (b) 48-h forecast. Locations of radiosonde stations are denoted by the “+” symbols. The solid black line denotes the location of the background cross section used to define anomalies.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3371.1
West–east cross sections of 24-h WRF forecast model bias of (a) potential temperature (contour interval 0.5 K), (b) water vapor mixing ratio (contour interval 0.02 g kg−1); and plan views of (c) 500-hPa geopotential height (contour interval 1 m), and (d) tropopause potential temperature (contour interval 0.25 K). In (a),(b), the solid black (dashed gray) line is the WRF (GFS) composite tropopause. In (c),(d), the heavy solid (dashed) contours correspond to the WRF (GFS) composite field, with a contour interval of 30 m in (c) and 5 K in (d). The 0 contour in the bias fields is denoted by the thin solid contour.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3371.1
Tropopause polar cyclone sample density from 2 Aug 2007 to 31 Jul 2009. Values are equal to the number of vortex samples within a 5° latitude × 15° longitude box divided by the cosine of latitude. Dashed lines correspond to the 20, 30, and 40 contours, while solid lines correspond to the 50, 60, 70, and 80 contours.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3371.1
(a) Skew T–logp diagram of composite temperature and dewpoint temperature at the vortex core (solid) and at the background grid point (dashed). (b) Vortex–background difference plots of temperature (solid) and relative humidity (dashed). The plots to the right show the corresponding PV in (a) and PV anomaly from the background in (b) in PVUs at the vortex core (solid) and background grid point (dashed). Plots are valid at the model analysis time.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3371.1
Skew T–logp diagram of composite temperature and dewpoint temperature at the vortex core (solid) and at the background grid point (dashed) for (a) winter (December–January–February), (b) spring (March–April–May), (c) summer (June–July–August), and (d) autumn (September–October–November). The plots to the right show the corresponding PV in PVUs at the vortex core (solid) and background grid point (dashed). Plots are valid at the model analysis time.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3371.1
Composite (a) tropopause potential temperature (contour interval is 2 K) and (b) 500-hPa geopotential height (contour interval 30 m) centered on the vortex core.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3371.1
Composite west–east cross-vortex sections of the anomalous (a) temperature (K), (b) υ-wind component (m s−1), (c) Ertel potential vorticity (PVU), and (d) relative humidity (%). The thick, solid (dashed) black contour is the composite (background) tropopause. The 0 contour is denoted by the thin, solid contour.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3371.1
Composite west–east cross-vortex sections of the (a) total diabatic heating rate, (b) total diabatic heating rate anomalies, (c) radiative heating rate, (d) radiative heating rate anomalies, (e) latent heating rate, and (f) latent heating rate anomalies. Units are in K day−1 and the contour interval is 0.3 K day−1 in (a),(c),(e) and 0.2 K day−1 in (b),(d),(f). The thick, solid (dashed) black contour is the composite (background) tropopause. The 0 contour is denoted by the thin, solid contour.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3371.1
Composite west–east cross-vortex sections of the (a) vertical motion (Pa s−1) separated into (b) symmetric and (c) asymmetric components. The color interval is 0.01 Pa s−1. The thick, solid (dashed) black contour is the composite (background) tropopause. The 0 contour is denoted by the thin, solid contour.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3371.1
Composite west–east cross-vortex sections of the anomalous EPV tendency due to (a) all diabatic effects, (b) radiation, (c) latent heating, and (d) all diabatic effects except for radiation and latent heating. Units are PVU day−1 and the contour interval is 0.04 PVU day−1. The thick, solid (dashed) black contour is the composite (background) tropopause. The 0 contour is denoted by the thin, solid contour.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3371.1
Composite tropopause potential temperature anomalies (K) in the (a) winter (DJF), (b) spring (MAM), (c) summer (JJA), and (d) autumn (SON) with a contour interval of 2 K. The 0 contour is denoted by the bold, solid contour. The total number of samples used in the vortex composite is 132 for DJF, 159 for MAM, 146 for JJA, and 131 for SON.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3371.1
Composite west–east cross-vortex sections of the anomalous (a),(b) temperature (K); (c),(d) relative humidity (%); and (e),(f) radiative heating rate (K day−1) in the (left) winter (DJF) and (right) summer (JJA). The contour interval is 1 K in (a),(b); 3% in (c),(d); and 0.2 K day−1 in (e),(f). The thick, solid (dashed) black contour is the composite (background) tropopause. The total number of samples used in the vortex composite for DJF is 132 and 146 for JJA. The 0 contour is denoted by the thin, solid contour.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3371.1
Composite west–east cross-vortex sections of the anomalous EPV tendency due to (a),(b) all diabatic effects; (c),(d) radiation; and (e),(f) latent heating in the (left) winter (DJF) and (right) summer (JJA). Units PVU day−1 and the contour interval is 0.04 PVU day−1. The thick, solid (dashed) black contour is the composite (background) tropopause. The total number of samples used in the vortex composite for DJF is 132 and 146 for JJA. The 0 contour is denoted by the thin, solid contour.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3371.1
The background forecast fields are averaged at fixed locations for every model simulation, regardless of whether a vortex is present in the domain.
