An overview is given of current issues concerning the coupling between the stratosphere and troposphere. The tropopause region, more generally the upper troposphere/lower stratosphere, is the region of direct contact where exchange of material takes place. Dynamical coupling through angular momentum transfer by waves occurs nonlocally, and provides a generally negative torque on the stratosphere which drives an equator-to-pole circulation (i.e., towards the Earth’s axis of rotation). This wave-driven circulation is the principal mechanism for intraseasonal and interannual variability in the extratropical stratosphere. Although such variability is generally dynamical in origin, there are important chemical and radiative feedbacks. The location of the tropopause has implications for radiative forcing of climate, through its effect on the distribution of relatively short-lived greenhouse gases (ozone and water vapour). Some outstanding puzzles in our current understanding are identified. Attention is focused on possible climate sensitivities, and how these may be tested and constrained. Results from the Canadian Middle Atmosphere Model (CMAM), a fully interactive radiative-chemical-dynamical general circulation model, are used to illustrate some of the points.
Improvements in our understanding of transport processes in the stratosphere have progressed hand in hand with advances in understanding of stratospheric dynamics and with accumulating remote and in situ observations of the distributions of, and relationships between, stratospheric tracers. It is convenient to regard the stratosphere as being separated into four regions: the summer hemisphere, the tropics, the wintertime midlatitude “surf zone”, and the winter polar vortex. Stratospheric transport is dominated by mean diabatic advection (upwelling in the tropics, downwelling in the surf zone and the vortex) and, especially, by rapid isentropic stirring within the surf zone. These characteristics determine the global-scale distributions of tracers, and their mutual relationships. Despite our much-improved understanding of these processes, many chemical transport models still appear to exhibit significant shortcomings in simulating stratospheric transport, as is evidenced by their tendency to underestimate the age of stratospheric air.
Interannual variation is a year-to-year variation which is defined as a deviation from the climatological annual cycle of a meteorological quantity. It can be caused by a variation of an external forcing of the atmospheric circulation system, or can be generated internally within the system. On the other hand, intraseasonal variation is a low-frequency variation within a season, and it is basically considered to be a result of internal processes which may exist even under constant external conditions. In this article, some observational facts on the intraseasonal and interannual variations of the polar stratosphere are presented, and the use of a hierarchy of numerical models to understand the stratospheric variations is reviewed systematically. Numerical models can be roughly divided into three classes based on their complexity; simple low-order models, medium mechanistic circulation models, and complex general circulation models. In order to understand the stratospheric variations, a hierarchy of stratosphere-only models have been used under an assumption of “slave stratospheric-variations” or “independent stratospheric-variations”. Parameter sweep experiment, in which many trials of computations in parameter space are done by sweeping the value of a control parameter, is a powerful method to understand complex behavior in the models. Recently, however, the importance of the coupled variability of the troposphere and the stratosphere was pointed out for the intraseasonal and interannual variations. In addition to numerical studies with simple or complex models, we have done some parameter sweep experiments with three-dimensional mechanistic circulation models to understand the troposphere-stratosphere coupled variability. All the effects of external forcings that might cause interannual variations can be excluded in the numerical experiments to focus only on internally generated variations within the coupled system. The obtained intraseasonal and interannual variations have some similar characteristics of the real atmosphere in some realistic parameter ranges. Roles of the interannual variations of the external forcings are discussed, which might be significant even if the amplitude is small.
Changes in water vapor and methane covering the period 1992 through April 2001 are discussed. Global increases in 2 × CH4 + H2O are evident, however there is significant spatial structure at levels between 10 and 100 hPa. Anti-correlated decreases in methane and increases in water vapor are noted in the upper stratosphere; these are associated with a slowing of the mass flux into the upper stratosphere and a concomitant increase in residence time above 10 hPa. Increases greater than those due to the surface increase in methane are more difficult to explain. One possibility is changes in the water vapor flux into the middle and upper stratosphere associated with a widening of the tropical upwelling. Another possible cause is a change in the ratio of summer/winter net upwelling. Additionally, the first 3 years of the period considered appear anomalous at many levels simultaneously. It is postulated that this is due to enhanced vertical mixing at the start of the data record.
Variability in planetary wave forcing from the troposphere to the stratosphere is reflected in changes in ozone transport, due to fluctuations in the stratospheric Brewer-Dobson circulation and eddy mixing. This work examines the space-time patterns of correlations between column ozone tendency and planetary wave Eliassen-Palm (EP) flux into the lower stratosphere, using monthly mean data for 1979-2000. Strong correlations are found during winter-spring in both hemispheres, with out-of-phase ozone changes between the tropics and middle-high latitudes. Springtime polar ozone is strongly influenced by wave forcing in both the Arctic and Antarctic. The ozone tendency-wave forcing correlations are combined with observed variations in EP flux to estimate the dynamic contribution to decadal-scale NH ozone trends. These calculations suggest that interannual changes in EP flux contribute∼20-30% of the observed trends in column ozone over 35-60°N during the past two decades.
A variety of climate forcings are now thought to be able to influence planetary wave dynamics in the troposphere by affecting the propagation of planetary waves out of the troposphere. However, this propagation pattern is sensitive to the details of the corresponding zonal wind changes. Here we discuss two forcing mechanisms that alter zonal winds and subsequent tropospheric responses: changes in atmospheric CO2 concentrations, and solar forcing in conjunction with the QBO. Increased atmospheric CO2 concentrations can be shown to influence planetary wave refraction so as to produce an intensified residual circulation in the subtropical lower stratosphere (which increases transport of tropospheric species into the stratosphere). In our GCM experiments, the low latitude response appears qualitatively robust over a wide range of tropical warming magnitudes, although the quantitative circulation change depends upon the degree of tropical warming as influenced by convection and cloud cover changes; it varies by a factor of three with a factor of three change in tropical warming. At higher latitudes, this equatorward planetary wave refraction has been associated with an increase in the high phase of the Arctic Oscillation. In the model experiments, the extratropical response depends upon the magnitude of both low and high latitude warming in the troposphere; with SST and sea ice changes that result in a weaker Hadley Cell and greater high latitude warming, the Arctic Oscillation phase change may be negative. The QBO alters the latitudinal gradient of the zonal wind in the stratosphere, and solar heating, in association with ozone response, alters the vertical gradient of the zonal wind. Both gradients affect the refractive properties of planetary waves uniquely for each individual combination of tropical east/west winds and solar maximum/minimum activity. In the model, when we consider solar maximum compared to solar minimum conditions, the east (west) phase of the QBO results in a relative high (low) phase of the Arctic Oscillation with corresponding temperature changes. Observed and modeled surface air temperature variations calculated between the solar cycle extremes in the different QBO phases are similar in magnitude to those derived from regression of monthly data on the AO, both being on the order of observed interannual variations.
Interannual variations of the general circulation and polar stratospheric ozone losses are investigated by using a general circulation model (GCM) developed at Kyushu University. The GCM includes simplified ozone photochemistry interactively coupled with radiation and dynamics in the GCM. Polar ozone depletion is brought about in the GCM by a parameterized ozone loss term. We performed an ‘ozone depletion experiment’ over 50 successive years with stratospheric ozone losses occurring over the Arctic and Antarctic polar regions. In addition, a 50-year ‘control experiment’ without such losses was also performed. Results of the ozone depletion experiment show large interannual variations of the general circulation and polar ozone losses, especially in the Northern Hemisphere winter and spring. It is found that stratosphere-troposphere coupled interannual variations are caused not only by dynamical conditions, e.g., strength of the polar vortex and planetary wave activities, but also by interaction mechanisms between dynamical and ozone depletion processes. The resultant interannual variability of the general circulation in the stratosphere becomes larger than that in the control experiment. Moreover, influences of the stratospheric polar night jet extend to the troposphere during late spring; overall threedimensional patterns of the interannual variations in dynamical fields seem to coincide well with those of the Arctic Oscillation. On the other hand, for the Southern Hemisphere, it is found that there exists a remarkable interaction between meridional transport of ozone controlled by polar night jets and UV heating in the polar lower stratosphere. Strong and cold polar vortices lead to less ozone abundance in the polar region, resulting in less UV heating and lower temperatures, along with strengthening of the polar vortices themselves. As a result, even in the control experiment, interannual variability of ozone and temperature fields in the polar lower stratosphere is comparable to that for the ozone depletion experiment.
We have conducted GPS radiosonde and ozonesonde observations on board the research vessel “Shoyo-Maru” in the equatorial eastern Pacific. These observations took place in September and October 1999 as a part of the Soundings of Ozone and Water in the Equatorial Region (SOWER)/Pacific mission. The mean profile of ozone is similar to that for the dry season (September to October) at San Cristóbal, Galápagos (0.9°S, 89.6°W) located in the equatorial eastern Pacific. The mean tropospheric ozone concentration is about 40 ppbv with a maximum in the mid-troposphere. Compared with the mean profile during the dry season at Watukosek, Indonesia (7.5°S, 112.6°E), this mid-tropospheric maximum is larger, and a sharp increase of ozone below the tropopause begins at a lower altitude for the Shoyo-Maru profile. We frequently observed layers in which ozone and humidity are highly anti-correlated. These layers have vertical scales from several kilometers to several hundred meters. Horizontal scales of these layers are roughly 1000 km, which may correspond to time scales of about 2 days, since the vessel sailed about 500 km/day. These layers are related to northerly winds, which bring in wet and ozone-poor air from the inter-tropical convergence zone situated in the northern side of the main cruise track. Similar layers were observed in ozone profiles at San Cristóbal and Watukosek, mostly during the dry season, suggesting the existence of layers advected without vertical mixing.
The tropical tropopause layer (TTL) is a transition region between the troposphere and the stratosphere. In this study the vertical extent of the TTL is diagnosed from radiosonde and ozonesonde profiles in the tropics and a climatology of this layer is presented. The radiative balance in the TTL is also characterized. The TTL is locally defined as extending from the level of the lapse rate minimum at 10-12 km to the cold point tropopause (CPT) at 16-17 km. The minimum in lapse rate represents the level of maximum convective impact on upper tropospheric temperatures, which is found to closely correspond to a minimum in ozone. Variations in this level are correlated with convective activity as measured by satellite brightness temperatures and Outgoing Longwave Radiation (OLR). At the cold point, the TTL height is nearly uniform throughout the tropics, and has a pronounced annual cycle. There are regional variations in the altitude of the lower boundary of the TTL. Interannual variations of the TTL result from changes in the large scale organization of convective activity, such as from the El-Niño Southern Oscillation (ENSO). Over the last 40 years, records indicate an increase (200-400 m) in the height of both the cold point tropopause and the level of minimum lapse rate. To better understand vertical transport in the TTL, the clear sky radiative heating rate is diagnosed using a sophisticated radiative transfer scheme. The level of zero radiative heating occurs roughly 1 km below the CPT, implying that convection needs to loft air 4-5 km above the base of the TTL if the air is to eventually enter the stratosphere.
We have analyzed small-scale fluctuations of temperature in the stratosphere using radio occultation data from the GPS/MET (GPS/Meteorology) experiment. From this, we have determined a vertical wave number spectrum of the normalized temperature fluctuations (T"/T) at 20-30 km and 30-40 km during three periods in June/July 1995, October 1995 and February 1997. The spectra at 20-30 km in the equatorial region in February 1997 agreed very well with radiosonde results in Indonesia as well as with a model spectrum assuming a linear saturation of gravity waves. We have found that the GPS radio occultation technique could measure meso-scale temperature perturbations in the lower stratosphere with a vertical wavelength down to 400 m. We have investigated the dependence of the mean wave number spectra on latitude, height and season. At 20-30 km the power law index for short wavelengths (<2 km)was about -3, consistent with a saturated gravity wave model. However, the spectral density wassometimes smaller than that predicted by the model. At 30-40 km both the spectral slope and densityagreed well with the model for wavelengths shorter than about 1.5 km, though the slope was moregradual for small m. We have calculated the variance (T"/T)² integrating the spectra in two verticalwavelength ranges: 10-2.5 km and 2-0.4 km, and estimated potential energy per unit mass (Ep). Seasonaland latitudinal variations of Ep were evident at 20-30 km, in particular, Ep was highly enhancednear the equator for both long and short wavelength ranges. At 30-40 km the enhancement of Ep at lowlatitudes became less evident than at 20-30 km.
An efficient method of solution is developed for the Warner and McIntyre parameterization of the drag associated with nonhydrostatic non-orographic inertia-gravity waves. The scheme is sufficiently fast as to enable, for the first time, fully interactive multi-year climate simulations that include the effects of rotation and nonhydrostatic wave dynamics in the parameterization of non-orographic gravity wave drag. It is found that the new scheme alleviates much of the middle-atmosphere wind biases that occur in the Canadian Middle Atmosphere Model when either of its two operational hydrostatic nonorographic gravity wave drag parameterizations are used. The addition of the nonhydrostatic process of back-reflection has an important impact on the amount momentum flux launched into the stratosphere by the parameterization scheme. This quantity is a free parameter in the problem and it is specified here to be independent of time and geographic location. However, due to back-reflection, the net momentum flux that actually enters the stratosphere undergoes a systematic seasonal and latitudinal variation which is a consequence of the seasonal and latitudinal variation of the winds and temperatures in the middle atmosphere. This strong influence results in a characteristic latitudinal distribution for the net momentum flux entering the stratosphere in winter and summer. During these seasons, mid- to high-latitude launch momentum flux that is directed oppositely to the mesospheric jet can be reduced by as much as 75% due to back-reflection. The momentum flux launched into the stratosphere at tropical latitudes, however, is relatively unaffected by backreflection. This has important implications for the forcing of tropical oscillations such as the semi-annual oscillation and the quasi-biennial oscillation in general circulation models.
The size of the changes of the stratospheric heights and temperatures which can be attributed to the 11-year sunspot cycle (SSC) is discussed with emphasis on the summer seasons. It is pointed out that the signal is strong and consistent during the northern summer, but weak during the southern summer, January/February. During this time of year the northern winters are dynamically very disturbed and the global signal of the SSC is modulated by the QBO. The study puts earlier work on a firmer ground and gives the community of modellers who simulate the solar signal with GCMs quantitative values for comparison.
Effect of the modulation of the Polar-night jet oscillation (PJO) in winter time by the 11-year solar cycle is examined by the observational data from 1979 to 1999. It is found that zonal wind and the E-P flux anomalies appear commonly in the subtropical upper stratosphere in early winter of both the Northern and Southern Hemispheres as a response to meridional UV heating contrast. These zonal wind anomalies are found to propagate poleward and downward with development as a seasonal march in both hemispheres. Although the length of the record is limited, it is suggested from the available data that the signal due to solar activity appears as the time evolution of the PJO triggered by solar forcing at early winter in both hemispheres. Differences in the signals between the Northern and Southern Hemispheres during late winter are explained in terms of the different characteristics of the PJO in each hemisphere. A significant temperature signal is also found to appear in the Southern Hemisphere in late winter under a solar maximum condition.
A composite analysis with respect to the phases of the equatorial quasi-biennial oscillation (QBO) is made for the zonal-mean zonal flow and the Eliassen-Palm (EP) flux during late winter to spring in the Southern Hemisphere. It is shown that deceleration in the lower stratosphere in September to October is more rapid in the easterly phase of the QBO than in the westerly phase, which makes the stratospheric zonal wind in November weaker. Such deceleration is consistent with the larger convergence of the EP flux observed in the easterly phase, which is associated with the stronger upward flux from the troposphere. It is also found that the equatorward flux in the upper troposphere is stronger in the westerly phase. These results for the EP flux diagnostics are discussed in terms of angular momentum budget of the atmosphere within a supposed zonally symmetric “body-of-rotation”. The composite difference of the upward flux is dominated by zonal wavenumber 1-3 component, whereas wavenumber 4-6 is predominant for the equatorward flux. Accordingly, it is suggested to be inappropriate to explain the difference of these fluxes between the QBO phases by considering the alternative directions of propagation of particular wave components due to change in refractive index. It seems that activities of wave disturbances in both scales are modified respectively by the QBO.
The interannual variability of the decay of lower stratospheric Arctic vortices is examined using NCEP/NCAR re-analyses between 1958 and 2000. There is large interannual variability in the characteristics of the decay of the vortex air, with very different characteristics for early and late vortex breakups. In early breakup years (when the vortex breaks up in February and early March) the remnants of the vortex survive as coherent potential vorticity structures for around two months, whereas in late breakups (late April and May) the potential vorticity remnants quickly disappear. There is a similar contrast in the stirring around the vortex between early and late breakup years, as diagnosed by the lengthening of material contours in contour advection calculations. In years with an early breakup there is a gradual decrease in the stretching rates from large winter to small summer values, whereas in late breakup years stretching rates are roughly constant until late spring when there is a rapid decrease. These differences in the decay of coherent vortex structures and stirring suggest that there are large differences in the mixing of vortex air into the surrounding middle latitudes between years with early and late breakups.
A statistical analysis is made of the interannual variability of stratospheric planetary waves in the Southern Hemisphere (SH) in late winter using NCEP/NCAR reanalysis data gathered over a 20 year period from 1979 to 1998. The study investigates the dynamical coupling between the planetary waves and the tropospheric circulation, by focusing on the activity of transient baroclinic disturbances. With the aid of EOF analysis applied to the 10 hPa monthly mean height anomalies and tropospheric circulation patterns, it is found that the year-to-year variation of quasi-steady planetary waves in late winter (September and October) is characterized by a longitudinal phase shift of zonal wavenumber 1 as well as by a variation in wave amplitudes. An eastward (westward) shift of the stratospheric planetary waves corresponds to a double-jet (single-jet) structure of the upper tropospheric (300 hPa) zonal winds. On the other hand, planetary wave amplitude variation is closely related to that of the meridional heat flux in the upper troposphere. These results are reinforced by the Eliassen-Palm flux diagnosis composed of the stratospheric EOF scores, and hence the effect of tropospheric transient waves on the forcing to the stratospheric planetary waves is stressed.
This study discovered an eastward-propagating circulation pattern in the stratosphere of the Southern Hemisphere based on the Met Office stratospheric assimilated data and the TOMS total ozone. This pattern is named the Stratospheric Antarctic Intraseasonal Oscillation (SAIO) because of its intraseasonal time scale and its repeated appearance in the high latitudes of the Southern Hemisphere. The SAIO exhibits a wave-number-one structure and propagates eastward around the globe in about 30 days, indicating a periodicity of 30 days. It is characterized by a deep vertical structure extending from the upper troposphere to the upper stratosphere with the amplitude increasing rapidly with height below 5 hPa and decreasing slowly with height above. The pattern exhibits a westward-tilting vertical structure with increasing height below 5 hPa and a more barotropic structure above. This westward-tilting feature is most evident in the eastern hemisphere. The close correspondence between the total ozone and circulation implies the impacts of the SAIO on the ozone hole. The SAIO exhibits active wave activity in the longitudinal band outside the jet stream, i.e., from 60°E eastward to 90°W, and tends to grow in the longitudinal band from 60°E to 120°E where energy is converted both barotropically an baroclinically from the mean flow to the SAIO. Outside this region, the SAIO tends to decay and feed energy back to the mean flow mostly through barotropic processes. The SAIO wave activity propagates upward from the upper troposphere to the upper stratosphere downstream of the high-rising topography in Antarctica. The feature is likely to be the manifestation of a topographically forced planetary wave. However, the periodicity cannot explained by the Rossby wave dispersion. Possible mechanisms are discussed.
Three-dimensional structure of quasi-stationary circulation anomalies observed in the course of a typical life cycle of an interannual seesaw between the surface Aleutian and Icelandic lows (AL and IL, respectively) is examined by using a reanalysis data set for the last three decades. A diagnosis is applied in a particular framework where the 31-day mean anomalies are regarded as stationary Rossby waves embedded in the zonally varying climatological-mean flow. It reveals that the upward propagation of wave activity into the stratosphere occurs in late winter primarily from the tropospheric anomalies corresponding to the anomalous IL, which develops below the entrance region of the lower-stratospheric polar night jet as a remote influence from the North Pacific. Accordingly, the North Atlantic anomalies exhibit apparent amplification and westward phase tilt with height, whereas the North Pacific anomalies are much more like the external mode. It appears that, in the presence of the zonal wavenumber 1 (k = 1) component of the climatological-mean planetary waves, the polar-night jet over the Pacific is shifted too far north to allow the stationary anomalies associated with the anomalous AL to propagate upward as stationary Rossby waves. Unlike the predominant signal of the Arctic Oscillation (or annular mode) that alters the intensity of the polar vortex, the seesaw modifies the stratospheric planetary-wave patterns in a modest but signifi-cant manner. In late winter when the AL is weaker than normal associated with a particular phase of the seesaw, the k = 1 component is masked by the enhanced zonal wavenumber 2 (k = 2) component embedded in the intensified polar-night jet. In contrast, the predominant k = 1 component embedded in the relatively weak polar-night jet dominates over the diminished k = 2 component in late winter, when the AL is stronger associated with the other phase of the seesaw. Our examination of seven late-winter major events of stratospheric sudden warming over the three recent decades suggests that the polarity of the AL-IL seesaw might set up a condition of which planetary wave component (k = 1 or k = 2) is more strongly involved in a late-winter warming event.
Effects of zonal variations in sea surface temperature (SST) on synoptic eddy activity are examined using an atmospheric general circulation model under the perpetual January and the aqua-planet conditions. In the presence of zonal variations in tropical SST, upper tropospheric zonal wind displays large zonal asymmetry and the storm track is located downstream of the maximum westerly speed. When only extratropical SST varies in the east-west direction, a storm track develops along the sharp meridional SST gradient despite a nearly zonally uniform westerly jet in the upper troposphere. The latter result suggests that zonal variations in upper tropospheric westerlies contribute to but are not necessary for zonally confined storm tracks.
Low-frequency variability in an idealized troposphere-stratosphere coupled system is investigated from a viewpoint of the annular variability. Our previous numerical experiment (Taguchi et al. 2001) with a simple global circulation model under a perpetual-winter condition is redone for a much longer period of 10000 model days. Amplitude of a sinusoidal surface topography of zonal wavenumber one, h0, is changed from 0 m to 1600 m in 15 runs as an experimental parameter, to examine the role of forced planetary waves in the annular variability. The parameter sweep experiment shows the ubiquity of the annular variability in the sense that the leading empirical orthogonal function exhibits a meridional seesaw pattern in the troposphere which has strong zonally symmetric component for all the runs. However, the numerical experiment also reveals qualitative change of the annular variability with the topographic amplitude in detailed spatial structure and temporal variability; the nature of the annular variability is classified into three regimes depending on h0: In the regime (i) of 0 m ≤ h0 ≤ 300 m, the annular variability consists almost only of zonally symmetric component and has long time scales. In the regimes (ii) of 400 m ≤ h0 ≤ 600 m and (iii) of 700 m ≤ h0 ≤ 1600 m, the annular variability exhibits zonally asymmetric component of zonal wavenumber one as well as zonally symmetric component. The temporal variability is characterized by small standard deviation and negative skewness of time series of the leading mode in the regime (ii), while the standard deviation is large and the skewness is close to zero in the regime (iii). The connection of the annular variability to the stratosphere is also different among the three regimes. The annular-variability signature substantially penetrates from the surface to the stratosphere with a time lag of about 20 days in the regime (i), while it is confined in the troposphere in the regime (ii). In the regime (iii), the annular variability leads to stratospheric sudden warmings with a time lag of about 30 days. The present results suggest that the Northern and Southern annular modes in the real atmosphere can be different from each other in these aspects in contrast to the strong similarity noted by Thompson and Wallace (2000).
In connection with the maintenance of blocking flows by the migrating synoptic disturbances, we examine the effectiveness of the eddy straining mechanism proposed by Shutts (1983) in an equivalent barotropic β-channel model. The model used here is identical to that in Haines and Marshall (1987) (hereafter referred to as HM) except with a channel twice as wide. This model possesses two stationary solutions accompanying isolated structures in prescribed uniform westerlies when a vorticity forcing associated with the analytical modon solution is assumed: “blocking solution” similar to the modon solution and “zonal flow solution” characterized by dominant zonal flows. The infinitesimal transient eddies which mimic synoptic disturbances are generated by a wavemaker forcing located far upstream of the diffluence associated with the basic flow prescribed by stable stationary solutions. The effectiveness of the eddy straining mechanism is evaluated from the resemblance between the basic flow and the second-order flow induced by the time-averaged eddy potential vorticity (PV) flux divergence. The distribution of the time-averaged PV divergence for the blocking solution in our model is the same as in HM in a sense that there is a PV north/south, divergence/convergence dipole upstream of the diffluence of the basic flow. However, the computed second-order flow has a quadruple structure which tends to shift the blocking dipole downstream instead of the dipole structure enforcing the blocking as shown in HM. On the other hand, the second-order flow tends to maintain a weak diffluence associated with the zonal solution. Thus, the effectiveness of the eddy straining mechanism depends on the basic flow. The second-order flow for the blocking solution is also drastically deformed by a negligible distortion of the PV divergence field associated with a small change in the property of imposed eddies. The effectiveness of the eddy straining mechanism depends also on the property of synoptic eddies. Finally, this study suggests that the PV north/south, divergence/convergence dipole pattern does not necessarily maintain the blocking dipole. This remark is especially important for the observational study to assess the role of synoptic disturbances in the maintenance of blocking flows.